6.0 ANALYSIS OF THE RESULTS - PDF document

6.0 ANALYSIS OF THE RESULTS ........................................................ a Convoy -33 6.2 Convoying ......................................................................... 6.3 Independent Speed Control Target Acquisition 6.5 Momentary Target Loss ......................................................... 49 ..................................................... 6.6 Driver 1 AND RECOMMENDATIONS Recommendation 1 Recommendation 2 -57 ....................................................... 7.2.3 Recommendation 3 58 ....................................................... 7.2.4 Recommendation 58 ....................................................... 7.2.5 Recommendation 5 58 ....................................................... 7.2.6 Recommendation 6 58 ....................................................... 7.2.7 Recommendation 7 58 .................... APPENDIX A -- SOFTWARE (in volume 11) .... ....... A-1 ........................ APPENDIX B . HARDWARE (in volume 11) .. ................. APPENDIX C DATA REQUIREMENTS 1 APPENDIX D -- REVISED TEST D-1 E A SAMPLE E-1 main thrust this work to advance the technology headway-control systems Army for efficient convoying military trucks. automotive industry the headway-control function in connection with intelligent cruise control products for both passenger and trucks. analytical and experimental results obtained in this study will used to evaluate the hypothesis that headway-control systems based upon current technology will aid drivers and improve the performance convoy operations. done was improving the performance driverJvehicle regard to collision avoidance and automatically providing the capability for following a leading system for controlling headway. study addresses a partial-control robotics, including collision avoidance, automation, and remote driving. also has implications with respect to mobility the sense that steadiness and level forward speed, automatic control headway, facilitates efficient movement vehicle convoys. hypothesized that a synergy exists between the Army's interest in automating functions, under the Advanced Land Combat & T thrust, and in developing headway-control systems assist truck and several issues significance are identified that promote automatic headway control, especially when implemented into a convoy operation: Convoys can move faster while maintaining a formation. Reduced risk Automatic maintenance minimize the limitations posed on operating in bad weather or under lighting conditions; Headway control can compensate for perception when driving using night- vision means; More attention of the hver other tasks; inadvertent lack attention occurs, the headway-control can provide a temporary compensation; Critical headway situations warnings can overall safety convoy operation the Army seeks primarily to unload driving functions and enhance the capabilities vehicle operations, truck manufacturers seek enhance the ease driving while improving military and civilian types systems have many design hypothesized that dual-use can be technology supporting headway control. Rdot-R Diagram the Development basic nature the headway-control process may dimensional space by range (R) and range-rate (Rdot) axes. diagram is particularly useful in understanding physical interpretations range and range-rate signals--either the form instantaneous (Rdot, points or (i.e., connected set points) that would take place over a period time. It should be noted that thls discussion regarding a headway-control system generic in nature, and apply only to the operation a convoy. Later, the peculiarities that pertain a military convoy are discussed. R (axis) -T I I I I I I 1 R Headway control the (R, Rdot) fundamental understanding headway-control systems associated with a graphical points, lines, and regions the Rdot-R example, the point (0, Rh) headway-control systems headway-control algorithms in addition, trajectories this diagram have unique properties, which are readily apparent properly interprets the implications � 0 and Rdot 0 on the change in R. a straight line trajectory parabolic trajectories are shown arrow heads associated these trajectories (lines) all illustrate 0 on the left side the diagram � 0 the right There is "upward motion" the right quadrant "downward motion" the left quadrant. the R (i.e., where Rdot 0) time, a trajectory must have 0. Points on the ideas presented in Figures illustrates the basic features a practical order headway various speeds the preceding vehicle, to let desired headway Rh = Vp-Th, where Th called the headway (this equation can represent rules such as car length for each mph," for example). the vehicle has a very responsive cruise control system such that V = Vc where Vc to the cruise control, then, except during a short transient period, a speed the equation Vc = Vp + (R - Rh)/T will suffice provide a system that acts like a first headway controller. Cruise control or speed headway control First order headway controller R-Rdot Pertinent Driving Situations. standard use headway control the road starts with a following vehicle approaching a preceding vehicle behind. Under sensor on following vehicle will detect the preceding range that is considerably longer range where headway control starts. Once switching line (described the equation TsRdot + R = Rh) is crossed, a headway-control algorithm (based upon the equation Vc) this vehicle to adjust its speed accordingly. extent to which the command speed (Vc) satisfied, depends maximum deceleration and acceleration available the inherent dynamics the vehicle). Figure instantaneous value (Rdot, R) The important aspects from short range are summarized choices of Rmin and Dm as well as Am and T and Th. A ,= driving situations the lead vehicle less constant velocity. However, slow down highway situation or as convoy. The autonomous control headway needs to to changes Vp, and hence Rh. to illustrate what would the preceding vehicle suddenly jump to a hypothetical situation driving situation in finite deceleration used to Nevertheless a step function Vp to illustrate which a intentionally left blank speeds, and vehicle response characteristics. Variables that present even in the simple vehicle following another were examined insofar they offered meaningful obtained under controlled proving grounds conditions. include the headway responses that the system maintains, a function items during testing. following steps were taken the course performing this task: operation as it pertains this project was studied together with TACOM, so establish a "baseline convoy" (see section 4.0- Operational Concepts). using different types sensors and sensing technologies various types particular blackout operation) was also considered. desired results (e.g. establishing headway in terms speed and limits on operational conditions) was defined and outlined. Measurements and performance measures required to evaluate system performance were also determined (see in section the results, and test plan data that needs be collected in order enable a proper analysis the results was determined and defined (see the data requirement in Appendix Appropriate test procedures were developed to evaluate the various performance measures, using ranges for speeds headways that were determined given the limitations the test track environment. A detailed test plan was formulated (see Appendix 3.1.2 Task 2 - Design Testbed the planned experiments for the test vehicle were outlined, the instrumentation needed to acquire and record the data was defined. complete instrument package was designed as to the data-acquisition needs; identified and ordered, and the installation effort was planned. See for a complete description testbed. following steps were performing this task: instrumentation required data per 1 was determined and listed. necessary instrumentation the instruments, borne the test Outfit the Truck During this task, the full complement equipment planned for the test vehicle was obtained and proper functioning first in the laboratory and then headway control a range final report this project. steps were taken the course performing this task: A demonstration of the testbed system was conducted, as to TACOM's personnel with a physical impression the research work. (The demonstration was held on October results, findings, and conclusions the tests and simulations are documented and are presented in this final report. Milestones and objectives and goals this study were achieved progressive work that followed prescribed milestones. Requirements prepared for instrumentation and data acquisition package. Plan ready for evaluative headway-control systems Instrumentation and data-acquisition package for installing in the vehicle. Truck outfitted for testing. Initial experiments completed. Demonstration exercise completed. Final report delivered. following schedule chart (see Figure illustrates the sequence and the timing various activities and sub-tasks) that performed during this project in order to achieve the milestones listed was intentionally left blank An essential fmt convoy can operate is its must be satisfied order to initiate system is on, and convoy operation can the leader. system for disturbances (e.g. gap variations.) system cannot automatic correction for the (e.g. gap increased beyond sensing manual correction the driver is required to bring convoy back operation (depressing the accelerator pedal, this case). Failing "break the leader and follower become two independent vehicles. operative loop described above. figure, the block entitled "initiate convoy" involves turning on the main ("convoy switch"), a target by the commences to the block entitled operation." Throughout convoy operation mode the convoy switch is target leader is being tracked and control inputs are provided the accelerator The disturbances and ensuing corrective actions are parts the control loop, and are discussed in the corrective action 1 operative loop Once the convoy, any pair leader/follower at a "convoy operation," the controller continuously automatically adjusts the headway. Such isconnect system's input accelerator1 brake I /use last speed command before\ target a cruise logic and control inputs rules regarding the operation the sensor or the position the driver through the accelerator andlor brake pedals takes precedence. truck will down according such control inputs regardless headway gap. Second, the convoy switch automatic speed take place. Operational Scenarios is a description possible scenarios that might take place during the operation the scenarios can involve several possible control inputs. response for each case is 4.2.2.1 Scenario 1 - the system (either when launching en route). The system is shut-down status. The command the system is given switching the convoy switch (CS) to possible scenarios can be associated with has been decelerating (acceleration brake pedal command used). The lead vehicle, other target, may may not present ahead. vehicle has been coasting down (e.g. joining the rear the convoy). brake pedal command inputs are being used. applied ? w Sensor has N A Figure 12. Scenarios releasing accelerator truck was slowing down using its brakes, just been released (CS is on). Three possible scenarios releasing the has been slowing down a convoy. That can occur either when the convoy is just the leader was broken. (2) The truck has been "convoying," for some reason, the gap became (e.g. downhill travel). driver decelerated increase headway. The truck been moving without being a convoy, for some reason the driver slowing down. logic depicted chart (Figure illustrates the response for the various possible brake pedal is released. intentionally left Sensor "enable" Switch I ~1~-~~1 Controller To4ue I Command Driver's parametric settings (e.g., headway time, - - -- - Figure 15. Convoy control implemented in testbed switch is the "master switch" the system: it must set to its "on" in order system to be operative. When the convoy switch the controller and its commands are transmitted using protocol to the engine. the engine is controlled strictly the throttle, and the Throughout the following discussion, is assumed that the convoy switch switch determines whether the data the sensor is being relayed to not. Its purpose the experiments required that the preceding vehicle system for a prescribed period which could be holding the (2) based the operational outlined in section sensor's data do not reach that there is no system keeps a constant speed, like a the required signals (either for performing control tasks or just data acquisition) were already available the truck, For example, the accelerator pedal position was readily available on the 5-1587 communication buss the electronically controlled engine, while the activation the brake had to connecting a special voltage sensor the brake-light circuit. Table sensors that were specially installed testbed, as the information provided them was the original M915A2. Table 2 (taken from Appendix C - "Data Requirement Plan") lists all data signals during the Range Rate ) from the sensor to a detected object distance from the sensor a detected ft fps V Ax Yr Csw VORAD Sensor VORAD Forward velocity the headway-controlled truck Forward acceleration the vehicle Yaw rate the vehicle Rotational position the steering wheel Brake Headway time g degke C deg % (- fps sec Speed transducer Yaw-rate transducer Cac Lbr Vset TH SAE data Brake light circuit display dial display dial Accelerator pedal position Boolean variable indicating brake pedal status: 0 brake pedal is brake pedal depressed speed value exceeded; set Desired headway Valid target Command speed Sensor engaged (-) f~s % (- Boolean variable to filter detected objects: 1 detected object b possibly adjust headway to 0 Velocity command signal from the headway control command signal Boolean variable indicating sensor data status: 0 sensor data is sensor data ksent to the The data from the sensors, the other signals listed a Macintosh PowerBook computer using shows the testbed and the instruments. Figure 17. Testbed instrumentation layout a schematic various items that data acquisition are composed of, linked between themselves and testbed truck. This page was intentionally lefr blank These tests assess the ability following vehicle pick up a preceding then to move into position a convoy a desired range. Figure illustrates the vehicle joins convoy from dynamics line. 3 4 5 6 7 8 9 10 Starting point (0) Joining a tests performed is desirable the driver turning on the headway control) that is the point (0, Rh). twofold. First, the sensor signals for Rdot and R and out long range. second, the convoy controller will cause the truck to initially to dynamics line, point that above the dynamics line (this is the usual for joining a convoy). the vehicle's (Rdot, R) coordinates from the to the dynamics line 7 9 10 16 17 18 20 21 22 Test Description (73.3 fps) Th=2.0 sec. from 50 (73.3 fps) 35 mph (51.3 fps), Th=2.0 sec. from 40 mph (58.7 fps) to 35 mph (51.3 fps), Th=2.0 sec. from 40 to 30 (44.0 fps), Th=2.0 sec. from 40 mph (58.7 fps) to 25 mph (36.7 fps), Th=2.0 sec. to 40 Th=1.5 sec. from 50 (73.3 fps) Th=1.5 sec. (44.0 fps), Th=1.5 sec. (58.7 fps) (36.7 fps), Th=1.5 sec. from 40 Th=1.5 sec. Test Plan Ref. 5 2.2.1 (# 5) ?j 2.2.1 (# 6) $ 2.2.1 (# 7) ?j 2.2.1 (# 8) 5 2.2.1 (# 9) 5 2.2.1 (# 9) ?j 2.2.1 (# 2) 5 2.2.1 (# 3) $ 2.2.1 (# 4) ?j 2.2.1 (# 4) Ideally, after acquired, the transition to modulating the throttle (see technical in section time history plot range between the during such ideal transition and the pertinent mathematical expression are portrayed equation shown, the time time (see The algorithm employed in T=30 seconds to reflect the limited acceleration capability. Accordingly, test results presented range, existing seconds after acquisition, as "OK R = Rh. (1 -e-fl) + b. e-fl Time (sec) time history on a lead vehicle 4 8 show that velocity difference (i.e., Rdot can result following vehicle coming desired to on these results, drivers should approach a preceding vehicle traveling no more than preceding vehicle. That is, Rdot should exceed approximately ftlsec joining a (Rh) Desired range 3 8 4 1 1 5 3 5 7 tests performed data collected I l'es mph (58.7 fps). Th=1.5 sec.. high Decel ..... fps) to (58.7 fps). Th=1.5 sec.. high Decel ..... fps) to Th=1.5 sec.. high Decel ..... mph (73.3 fps) (58.7 fps). Th=1.5 sec.. low Decel ...... fps) to Th=1.5 sec.. low Decel ...... mph (73.3 fps) to mph (58.7 fps). Th=1.5 sec.. low Decel ...... from 50 mph (73.3 fps) to 40 mph (58.7 fps). Th=1.5 sec.. low Decel ...... from 50 mph (73.3 fps) to 40 mph (58.7 fps). Th=1.5 sec.. low Decel ...... mph (73.3 fps) to 35 mph Th=1.5 sec.. high Decel ..... from 50 mph (73.3 fps) to 35 mph (51.3 fps). Th=1.5 sec.. low Decel ...... mph (73.3 fps) to 35 mph (51.3 fps). Th=1.5 sec.. low Decel ...... from 50 mph (73.3 fps) to 35 (51.3 fps). Th=1.5 sec.. low Decel ...... from 50 mph (73.3 fps) to 35 mph (51.3 fps). Th=1.5 sec.. low Decel ...... (73.3 fps) 35 mph (51.3 fps). Th=1.5 sec.. low Decel ...... fps) to Th=1.5 sec.. high Accel ..... (73.3 fps). Th=1.5 sec.. low Accel ...... fps) to Th=1.5 sec.. low Accel ...... fps) to 30 Th=1.5 sec.. high Decel ..... (58.7 fps) to 30 Th=1.5 sec.. low Decel ...... from 30 mph (44.0 fps) to 50 mph (73.3 fps). Th=1.5 sec.. high Accel ..... 30 mph (44.0 fps) (73.3 fps). Th=1.5 sec.. high Accel ..... (73.3 fps). Th=1.5 sec.. low Accel ...... (44.0 fps) Th=1.5 sec.. low Accel ...... (44.0 fps) 25 mph Th=1.5 sec.. low Decel ...... to 25 (36.7 fps). Th=1.5 sec.. low Decel ...... from 25 mph (73.3 fps). Th=1.5 sec.. low Accel ...... 25 mph Th=1.5 sec.. low Accel ...... mph (73.3 Th=2.0 sec.. high Decel ..... from 50 mph (73.3 fps) to 40 mph (58.7 fps). Th=2.0 sec.. high Decel ..... ..... from 50 mph (73.3 fps) to 35 mph (5 1 . 3 fps). Thd.0 sec.. high Decel (44.0 fps). Thz2.0 sec.. high Decel ..... mph (58.7 (-14 0 fps) . Th=2.0 sec.. low Decel f'ps) . Th=2.0 sec.. low Accel (58 7 fps). Thd.0 sec., to 25 fps). Th=2.0 sec., to 25 Th=2.0 sec., low Decel Th=2.0 sec., low Accel mph (58.7 ips). Th=2.0 sec., low Accel Th=2.0 sec., low Accel (44.0 fps). Th=2.0 sec., low Accel 'est Plan g 2.2.2 (# 1) 5 2.2.2 (# 1) g 2.2.2 (# 1) g 2.2.2 (# 2) g 2.2.2 (# 2) 5 2.2.2 (# 2) g 2.2.2 (# 2) g 2.2.2 (# 2) g 2.2.2 (# 3) g 2.2.2 (# 4) g 2.2.2 (# 4) g 2.2.2 (# 4) g 2.2.2 (# 4) g 2.2.2 (# 4) g 2.2.2 (# 5) 2.2.2 (# 6) fj 2.2.2 (# 6) g 2.2.2 (# 7) 5 2.2.2 (# 8) g 2.2.2 (# 9) g 2.2.2 (# 9) g 2.2.2 (# 10) 5 2.2.2 (# 11) g 2.2.2 (# 13) g 2.2.2 (# 13) $ 2.2.2 (# 14) g 2.2.2 (# 15) 8 2.2.2 (# 17) 5 2.2.2 (# 17) 8 2.2.2 (# 18) $ 2.2.2 (# 19) p 2.2.2 (# 20) g 2.2.2 (# 21) g 2.2.2 (# 21) g 2.2.2 (# 22) 5 2.2.2 (# 23) $ 2.2.2 (# 24) $ 2.2.2 (# 24) g 2.2.2 (# 24) 2.2.2 (# 25) 0. I I I I I I I 10 15 nurrber 40 1 o.ge-. 0,9-. time range rate f lfps during following XY x I ,-.,-.a 1 I I + ,-. ?xx : , X :x X �: ' x, ..... . ....... ....... .,. ........:......... .:. .X. '. .X.. ....................... .:. - , X x : x: Xx: : x X ........ ..... ......... ....... ........ ......... .................. ' X .x .x.. 1 I.. .:. .:. : - x : : X : x: Similar to the range rate, which should be following, the deviation the actual range from the desired range should also zero during following. Figure shows the range error (the deviation range from desired headway range, average deviation) for following sessions. Figure depicts the same data a histogram. the deviation positive (smaller be due to delays and lags associated with the operation .......... !.........1.....................'................... - Average deviation (Rh-R) during following O.&. S .- - # o,,e-. Itb ,x, Y ....... .:. ........ .: ........ .:. .-..... .I.X.. ........................... : ........ - .... ....... .:. ..-...... ;.. ....... .;. ....... .k.. ....... x.. ....... .:. ........ .I .X.. - 8 x, ................................ . 0,6C ~eight& average ¢ of tin;: 091 ! .......... - ..................... the lead vehicle that in the AVp - ap plane there should a boundary between braking braking. Such a boundary will separate between those speed deceleration levels that will necessitate braking the driver and those that the be capable that boundary "VP (fps) Figure 27. Braking / no-bralung boundary when the on the testbed using parametric values that represent the driving conditions during values were used in solving AVp: the following truck's maximum deceleration (Dm) is 0.05g a 0.08 sec. delay (@) the instant to decelerate is introduced, and the instant truck develops maximum deceleration minimum range dnver to tolerate in the process speed adjustment ft. (Rmin) range gets shorter, driver perceives ft, driver applies the brakes) this value pertains to the particular driver it might vary with drivers) (Th) was alternately 1.5 seconds and 2.0 seconds. these calculations are represented VpO" Figures 29 and 30. These lines depict brake / boundaries for various values VpO according to equation (4). data from the tests (see Figure have been Figures 29 to compare with the lines VpO from equation Comparisons indicate that equation (4) is a reasonable predictor for the of 1.5 AVp (mph) Figure 29. Braking / seconds headway time A~ (gts) a,,, 4.05 0 4 30 lines of VP 0 (mph) 35 1. -40. - -- - -- - 1 - - -_ - - - 4.5 m + A (49, A(soj 0f49j * (49) Brakes were applied 0 Brakes were not applied parenthesis indicate VpO for tiat data p)lnc -20 40 Independent Speed This test is described 2.2.3 plan. It a total five trials, each involves maintaining different speed: 20,30,40,50, to examine maintaining a constant speed (cruise control operation). The results deviation from the desired speed and the standard deviation are depicted in Table 7. average, the truck fps which is practically (see section manner to maintain a constant speed. Sudden Target Acquisition Independent speed control deviation from a constant speed This test is described simulates a situation, or target that close for orderly speed adjustment joining a sixteen trials planned: eight cut-in and two headway the prescribed cut-in perform that lead vehicle changes its speed from mph, and the truck attempts follow this maneuver while maintaining a for which data was fifteen Standard deviation Nominal constant speed 20 Average speed 1 0.013 Range 3t 1;; ;;;; ;;;;;:;;;;;:;";;';~ -24 -20 d $4 32. Deceleration boundary that calls for driver's intervention, Th=2.0 sec. parabolic lines, which as deceleration boundaries in these figures, are written the deceleration is associated with them: Observing is evident that whenever the designed warning line (the parabola that required deceleration 0. which also when the warning light should be always resulted There were false alarms. However, the results further show that there were the driver intervened before the warning light (when the (Rdot, the warning parabola). This opposite to false alarm the driver should the required deceleration to avoid crash is less 0. Such amended warning lines are: 0.077g (for 1.5 sec. headway time), and 0.088g (for 2.0 sec. since the driving conditions during its boundaries, more conservative warning line might for normal might indeed trigger false alarms (unnecessary belated alarms 0.05g headway times) will appropriately serve this purpose. 6.6 Driver 1 Rather than relating to single set prescribed tests, this element the data goal is to identify those sets which the takes control the system initiates braking. Using the depicts the first instants brake application. that was made indicates 0. approximately the warning boundary for agrees with deceleration level controller design. deceleration limit 0.05g. that are with these deceleration values are shown 33. Brakc act~\,ation points Figure 34 depicts Dreq (required decelcratlon to a\,o~d a crash, brake application. Sincc on1 72 prcent the data points are within the f standard deviation. DrcLl considered as the most consistent indicator the necessity warning or for ;I niorc drahtlc crash-prevention control to crash we Table 2 in appendix C): Tc, shown pertain first brakes were applied. the data point+ arc the range f one standard Figure 36 depicts Ta (available reaction time, see Table in appendix at the instant the brakes were applied. the data points are included within the f one standard deviation. Being so inconsistent, this parameter appears to be a rather poor indication for warning. so far off the the situation, that being ignored. Similarly, Thm (actual headway time, see Table in appendix 37), and probably cannot used as reliable indication for a Dreq, Tc, Ta, and Thm show that them alone appears satisfactory for setting driver / general. However, a boundary based Figure 33, range and range-rate when the driver needs acting autonomously 0.05g line in Figure 33. 1.4 1.2 h 1- U 0, - 0.8 Figure 36. Ta activation points sec. S.D. = 0.4376 - (52% of the po~nts are w~thln Ave.tS.D.) X Ave,+S.D. X-% - - - - - - - - - - - - - - - - - - - - - X X X x X 'Aye. X 0.6 0.4 - X X X X - - X X Aye. -S.D, - - - - - - - - - - - - - - - - - - - - - - - - - X X X X 0.2 I I I 0 5 I0 15 20 25 Test No intentionally left intervened and applied the however, that a more conservative a warning threshold would be safer. involving independent speed control show that the speed performs satisfactory manner to maintain constant speed. (The average the system is 0.027 fps.) provides reliable range range-rate readings convoy control system is to use and to be same time. maintains convoy integrity when operated autonomously, or functionality standpoint, its operational algorithm, which was is comprised preceding vehicle is present, automatically attempt achieve and driver does takes precedence preceding vehicle ib present, and the driver does control input, the system la41 input. During the experiments, "operational philosophy" was found to function well. The following features were consistently pushes either thc hrakc pedal or accelerator pedal, the authority over systeni. There i\ conflict between these means input and the convoy control s!,stcrn. the driver releases either t hc hrllkc pedal the accelerator pedal that authority returned to the control system, thc conlmanded speed from the the speed the truck the time the pedal w~t rclcased (if a preceding vehicle is achicvc the desired hcadua! (if preceding vehicle software and hardware, thC protot!.pc \ystem, constitute a system for headway control. detailed description ol ttlc hard\\.;rrc IS and the software is described detail (actual code I In ;rppcnci~x are aimed ( I ) f'unhcr cnhanc~ng knowledge concerning I cxplorlnf from findings a string vehicles should investigated experimentally. During this study tests with just one following vehicle. Longer least four five vehicles convoy will allow real-life evaluation the impact that the stability phenomenon might have convoy operation. It is also to study nonidentical vehicles. 7.2.4 Recommendation 4 prototype headway-control system installed M915-A2 should receive further operational testing and development. The performance boundaries the prototype testbed, the test results and findings (listed are likely with military Convoy performance capabilities should probably be allow the larger headway gaps (perhaps even out higher relative speed higher deceleration for crash avoidance more severe speed-change maneuvers the lead vehicle more accurate 7.2.5 Recommendation 5 Provision for changing speed throughout convoy should be investigated. communication between cooperative convoy operation can be obtained. This approach might eliminate the need deceleration under certain the potential longitudinal stability Recommendation 6 relationships between various scenarios and the acceleration capability should investigated. Tests analyses should performed to speed as vehicle mass and 7.2.7 Recommendation 7 Human factors studies should performed and its associated automatic convoy control system largely handles the longitudinal control the vehicle with the advantages that (1) the convoying operation performed with greater diligence and precision what drivers are likely driverlsoldier is more capable performing other associated with the mission In order to provide the overall flexibility needed to problems that arise, the driver is able to override the the control system automatic measures included driver's observational capabilities are backup for the in situations where target, loses target, or indicates a false target. (In convoy application, the is the preceding vehicle.) Temporary losses target may automatically holding speed constant the sensor. this automatic countermeasure quickly resolve need to intervene convoy operation. However, correcting for temporary losses target could burden the driver the extent that the system would limited value. The goal to keep driver interventions to minimum number cases involving easily performed corrective actions. relationship between automatic convoy operation actions aimed establishing or re- establishing automatic convoy operation illustrated Figure 1. be satisfied automatic convoy operation: the system used performing the convoying function turned on and 13) vehicle presents suitable target the sensor. the driver brings hishcr vehicle into sensor range preceding vehicle. When conlroy ing system target, automatic convoy operation During automatic convoy operation. thc control system corrects errors in headway gap andlor vehicle speed. automatic correction for the situation (e.g., increased beyond sensing is required to bring the back into automatic operation. Failing to do so will break the this happens, the original follower bcconlc rhc Icadcr and the originally preceding vehicle will become vchiclc In a shoncned the original convoy. F 2 OUTSIDE WORLD P-. I DRIVER 1 Driver Q~mmands* I I Driver perceptions of *Driver command (1) driver's intention to accelerate, brake, the driver a 0 4 "--t-t INTELLIGENT VEHICLE (irtclrrding lzardware and associated software, control algorithms, etc.) which may be referred to auxiliary variables, will provide additional information be used the performance the convoy control system. pertaining to: sensor, (2) the vehicle, (3) the the controller will be used evaluate convoy The data pertain to sources is discussed sensor measures and range-rate (dR/dt) data pertaining is fundamental to evaluating and controlling headway. range values measured is and range-rates ftfs to + 20 ft/s (k 14 mph, or + 6 rnls). desired accuracy for range data f 2 m, and + 0.15 m/s for range-rate the precision these measurements depend upon the accuracy EatonNORAD the longitudinal direction, the essential vehicle data are velocity and acceleration. measure forward over a range from ft/s (15 mph, 7 m/s) to 110 ft/s (75 mph, 33 ds). desired accuracy is one percent. may need fifth wheel for this measurement velocity is available within communication-buss system the vehicle or instrumented with accelerometer for direct measurement Available sensors are 2 g range with resolution 0.01 g. use velocity and/or range-rate changes over measured periods time to calculate decelerations especially identify when the vehicle is turn? yaw rate and be measured. information is identifying whether rate is to be measured over a range degls to + 20 degls be measured over the range 150" to + 150" with a resolution of two degrees. Appendix C Computed Data addition to the data collected during the tests, there are auxiliary variables that we wish and evaluate. These auxiliary variables be derived the acquired data listed in and will the original data files. these variables enhance data providing additional information concerning operating within, and various control regimes such as those discussed earlier better understanding operating patterns also be achieved. Table lists the auxiliary variables that we plan later use. Available reaction time (for Rdot headway time to collision Target headway range the preceding vehicle Required Deceleration Ta = - Units sec 1 b=& (for Rdot 0) I 2 .R I I Measured command velocity R -Rh V,, = vp +- T fps R / sec Thrn = v P 1 R Tc = - - Rdot Rh = Th.Vp Vp = Rdot + V sec ft ~PS ft/sec/sec Appendix D 1 .3.1 test track is portrayed together with its pertinent curves are superelevated to support lateral acceleration, and the spiral sections provide a smooth transition between the straightways and the test track limited field +2 degrees sideways +3 headway-control system, that the sensor is able the target. Under conditions, no involuntary opposed to target should occur. the experiments that are required to evaluate convoy performance impacts, ideally involve long, straight test track with relatively sharp curves and rather short straightways. constraints imposed the geometry the test track have been carefully accounted for preparing the test control regimes encountered these operating situations are: free travel with target in range", (2) target no desire to control headway, and preceding vehicle. operating situation be investigated specially devised test procedure for evaluating convoy performance carefully defined scenario. Test Procedure section provides details concerning individual experiments that comprise proving-ground test Joining a Scenario 1: arbitrary speed, is joining the a convoy. That truck could represent either individual entity, vehicles. This situation entails preceding vehicle, executing speed adaptation.. Hardware requirements: in\tn~nlcnted M9 15-A2. vehicle with cruise control. radio comrnunicatlon betu~ecn Procedure: Using its cruise control. will be driven speed per belou . The test behind, driven the appropriate constant spced ( see table 1 Effort should be made prior to the actual detcmiine the relative initial vehicles along the track. 50 tha~ the +peed mostly done on the straightwa),. The dr~\,cr\ 5hould familiarize themselves with the the cur\'ccl ponlon 01' the track. Duration: The experiment shoulcl ht. perfonlied successfully once for each tZ \i~ccessSul will be the experimenter the tollowing guidelines: (1) executing speed exceptional events. maintaining the final speed +3 reaching and maintaining prescribed duration (see table Each speed change after the required period initial steady-state following is completed. steady state the final speed is obtained, should be maintained for approximately minutes. Table below lists initial speeds, speed changes, time duration maintaining steady-state, and approximated rates for speed changes. Test matrix for convoying 1 1 steady Final steady Speed change 1 speed / Desired/ speed 1 Desiredl Acceleration Duration Headway (sec) time �(set experiment should perform successfully once for each speed. successful termination be determined the experimenter using the following guidelines: (1) reaching and maintaining the prescribed speeds within approximately f 1 mph, (2) fulfilling the approximately two- minutes requirement for maintaining steady state speed. Special instructions: Sudden target truck that is moving under either speed or headway control needs to respond to preceding vehicle) that appears suddenly. In the headway control and automatic convoying, that the target appears orderly speed adaptation. situation necessitates more abrupt response than that for Distinction is also made sudden targets that are slower test vehicle, those that are faster. Hardware requirements: M915-A2, control, radio communication between the experiment starts straightway section the track, with the and the moving one behind the stabilized speeds per table (using cruise control). During this the experiment, mode. The unique testing even though taken. When between the vehicles is at the value, the experimenter switches the headway system from mode, thus obtaining suddenly acquired target. accuracies for parameters cut-in range the point 55 ft, speed of vehicles is approximately +1 times: eight cut- speeds, and two time settings. successful termination Momentary target loss truck that moving under convoy, needs to respond to momentary target loss. the context headway control and automatic convoying, "momentary target means that been steadily followed, disappears for period, and again. period, change speed is either minor and state, or does take place target loss convoy goes through relatively sharp some environmental conditions the preceding vehicle becomes temporarily invisible perhaps the shape and might its detection by sensor. This experiment addresses the the system operation during questionable Hardware requirements: Instrumented M9 15-A2, vehicle with cruise control, radio communication between Procedure: The experimental staning point I\ at the beginning of the straightway section of the track. with thc tnto vehicles moving as sped. The speed of the convoy will be determined by the cruise control the lead vehicle. However, the cruise "set speed" set to rhc experimenter switches the headway system modc a4 if'thc was momentary lost, then after prescribed period. rhc 4!*stem back to normal as the target rc-acqu~rcci. Thrcc convoy will be 20 rnph. I.or cactl \peed, three "target loss periods" 3 second\. h scconci4. procedure should hc pert'ornicd t~.icc: convoy operated at 1.5 once t'or 7 seconds (see table 4). Appendix E Velocity - fps 70 Time - sec Figure E-1. Actual speed and commanded speed when convoy slows down Range rate Range versus range rate Appendix E Forward acceleration - g 4x 1 o-~ Time - sec Longitudinal acceleration Torque command % 0 20 40 60 80 100 120 Time - sec the engine