CN-121990021-A - Head space optimal control method for virtual marshalling train following operation

CN121990021ACN 121990021 ACN121990021 ACN 121990021ACN-121990021-A

Abstract

The invention discloses a locomotive spacing optimization control method for virtual marshalling train following operation, which comprises the steps of obtaining and processing pilot vehicle state data, constructing a variable-domain fuzzy PID control algorithm based on virtual marshalling train following, designing an operation scene of the virtual marshalling train, processing and optimizing performance evaluation of speed and displacement of the virtual marshalling train algorithm following operation, and judging whether to finish according to time step. According to the invention, the optimal safe locomotive spacing between trains can be calculated according to the real-time running state and scheduling requirement of the trains, and the actual running locomotive spacing can be dynamically adjusted, so that the running safety and the running efficiency of the trains are improved.

Inventors

LU DEBIAO
WANG ZHONGLI
ZHANG TIANBO
LIANG YITING
CAI BOGEN
WANG JIAN
Shangguan wei
LIU JIANG
JIANG WEI
BA XIAOHUI
CHAI LINGUO

Assignees

北京交通大学

Dates

Publication Date: 20260508
Application Date: 20260114

Claims (10)

1. The locomotive spacing optimization control method for the follow-up operation of the virtual marshalling train is characterized by comprising the following steps: S1, acquiring and processing state data of a pilot vehicle, namely acquiring the state data of the position and the speed of the pilot vehicle and the line environment data of the line gradient and the curvature, synchronizing the state data and the line environment data to form a standardized data set, and providing basic input for the construction and the control execution of a follow-up algorithm; S2, constructing a variable domain fuzzy PID control algorithm based on virtual marshalling train following, namely based on the pilot vehicle state data and the line environment data output by S1, taking a train speed error and a position error of the pilot vehicle as input variables, inputting the input variables into a calculation module, constructing a variable domain fuzzy rule by the calculation module according to the input parameters, setting a telescopic factor on the input domain and a telescopic factor on the output domain, and calculating control parameters , , ; S3, designing an operation scene of the virtual marshalling train, namely independently configuring operation environment parameters of the virtual marshalling train as a preset operation environment of a subsequent control flow, wherein the operation environment parameters comprise an initial distance between a pilot train and a following train, a target operation speed interval, a line type and a marshalling state interval, and the line type comprises a straight section, a curve and a ramp; s4, speed and displacement processing of the following operation of the virtual marshalling train algorithm, wherein the speed and displacement processing is based on the real-time state data of S1 and the control parameters output by S2 , , Calculating the target speed of the following vehicle through an algorithm, and determining the minimum vehicle head distance by combining a safety braking model to realize dynamic following control; S5, optimizing performance evaluation, namely based on the difference value of the speed output by S4 and the target speed and the difference value data of the following distance and the target distance, comparing the control precision and stability of the variable domain fuzzy PID and the traditional algorithm, and outputting performance optimization indexes; And S6, judging whether the process is finished according to the time step, namely judging whether the process reaches a preset operation time or a performance standard threshold by taking the sensor sampling step of the S1 as a basic period and combining the performance evaluation result of the S5, returning to the S1 to acquire the latest data again if the process is not standard, repeatedly executing the control flow of the S2-S5, and stopping the control and outputting a final optimization result if the process is standard.
2. The method for optimizing and controlling the inter-locomotive distance of a virtual marshalling train following operation according to claim 1, wherein in step S1, specifically comprising: and acquiring dynamic state data of the pilot vehicle through various sensors, and transmitting the dynamic turntable data to a calculation module for processing, wherein the dynamic state data of the pilot vehicle comprises the current position, speed, acceleration and expected running path information of the pilot vehicle.
3. The method for optimizing and controlling the inter-locomotive distance of a virtual marshalling train following operation according to claim 1, wherein in step S2, specifically comprising: According to the speed error between the pilot vehicle and the following vehicle and the position error fed back by the controller, dynamically adjusting the speed control parameters of the following vehicle by a variable domain fuzzy PID control algorithm so as to adapt to different running scenes, wherein the method comprises the following steps of: s21, calculating input variables, wherein the method specifically comprises the following steps: Is arranged at The following speed in the time virtual marshalling operation is The actual displacement is The pilot vehicle speed is The actual displacement is By means of Speed of following vehicle at moment The pilot vehicle at the moment can calculate the speed error The specific expression is: ; the change in speed error is The specific expression is: ; The position error fed back by the controller is the actual displacement error of the following car and the pilot car in the virtual marshalling at the previous moment The specific expression is: ; ; Wherein, the For the set position interval between the following vehicle and the pilot vehicle in the virtual marshalling at the moment 0, the specific expression of the actual displacement error fed back by the controller is as follows: 。
4. the optimized control method for the inter-locomotive distance of a virtual marshalling train following operation according to claim 3, wherein step S2 further comprises: S22, establishing a fuzzy subset, which specifically comprises the following steps: The input to the variable domain fuzzy PID control algorithm is the speed error of the follower in the virtually grouped train Variation of speed error The output being part of PID control , , Setting up The domain of theory of (2) is , The domain of theory of (2) is , 、、 Are all in the domain of For each input variable, the fuzzy subsets are 7, and are respectively negative large # ) Negative middle% ) Negative small% ) Zero% ) Just small% ) Middle% ) Zhengdazhang (Chinese character of 'zhengdazhi') )。
5. The method for optimizing and controlling the inter-locomotive distance of a virtual marshalling train following operation according to claim 4, wherein step S2 further comprises: S23, setting fuzzy rules, specifically comprising adopting " The fuzzy rule of the form judges output variables according to the input variables, and the specific expression is as follows: ; the paste rule has two output variables, so Judging the output variable by the fuzzy rule, wherein the specific rule setting is judged according to experience: a) When (when) Corresponding to the fuzzy subset PB/NB, Or (b) In the time-course of which the first and second contact surfaces, Take large value to raise corresponding rapidity, in order to prevent The instantaneous value is too large and, Small value should be taken, and integration is limited to avoid large overshoot ; B) When (when) Corresponding to the fuzzy subset PM/NM, Or (b) In the time-course of which the first and second contact surfaces, Taking a small value, making the system overshoot small accordingly, in which case, The influence on the system response is large, and a small value is required; Is increased appropriately to improve system stability; c) When (when) Corresponding to the fuzzy subset PS/NS/ZO, 、 Or (b) In the time-course of which the first and second contact surfaces, 、 The proper increase leads the system to have good stability, if at the moment Smaller, suitably medium sized Avoiding oscillation of the system if at this time The size of the particles is larger than the size of the particles, Taking small values.
6. The optimized control method for the inter-locomotive distance of a virtual marshalling train following operation according to claim 5, wherein step S2 further comprises: S24, setting variable domain expansion factors, namely setting different expansion factors by combining nonlinear environmental resistance of line gradient and curvature radius existing in a line scene, and dynamically adjusting the domain range of an input variable And domain range of output variables Controlling the train; Setting the sampling step length of the controller The input variable is recorded as The scale factor on the input discourse domain is recorded as The output variable is recorded as The scale factor on the output universe is recorded as , And Take the following forms: ; ; Wherein, the , , The value of (2) is related to the resistance factor of the line environment, when the curvature radius and the gradient value exist in the line environment, Taking 0.6, when the curvature radius and gradient value do not exist in the line environment, Taking 0.8; And S25, defuzzifying, namely defuzzifying by adopting an area barycenter method, taking the barycenter of the area surrounded by all membership function curves obtained by fuzzy reasoning and the abscissa as a final output value, wherein the formula is as follows: ; Wherein, the Is the first Membership of the fuzzy subset; Is the first The center of the individual fuzzy subsets, the center position or "centroid" of the membership function; is the number of fuzzy subsets.
7. The locomotive spacing optimization control method for the follow-up operation of the virtual marshalling trains according to claim 1 is characterized in that in step S3, a scene is designed to verify and optimize the variable-domain fuzzy PID control algorithm constructed in step S2, the scene design considers the operation state of the virtual marshalling trains, the initial position, the speed and the expected destination of the trains, and simulates various complex situations of train spacing control, the algorithm performance is further optimized through simulation of the operation scenes, and stability and high efficiency of the algorithm under different situations are ensured, wherein the complex situations comprise train acceleration, deceleration and emergency braking.
8. The method for optimizing and controlling the locomotive spacing of a virtual marshalling train to follow operation according to claim 1, wherein in step S4, the operation speed of the following train is dynamically adjusted according to real-time data feedback of a pilot train, and the speed and the locomotive spacing of the virtual marshalling train to be operated are generated in real time, specifically comprising: S41, acquiring and processing following vehicle speed data, wherein the following vehicle speed data specifically comprises the following vehicle speed is calculated through algorithm control and is set at The speed of the time following vehicle is The speed of the pilot vehicle is The speeds of the following vehicles and the pilot vehicles meet the following conditions: 。
9. The method for optimizing and controlling the inter-locomotive distance of a virtual marshalling train following operation according to claim 8, wherein in step S4, further comprising: s42, calculating the minimum safety distance of the train, wherein the minimum safety distance between the pilot vehicle and the following vehicle is calculated by utilizing a physical model and a safety algorithm based on real-time dynamic data of the pilot vehicle, and the pilot vehicle in virtual marshalling operation is set The displacement at the moment is The safety braking distance calculated according to the safety braking model is Adjacent follower vehicles behind the pilot vehicle are on The displacement at the moment is The safety braking distance calculated according to the safety braking model is The length of the train body is In order to avoid collisions of the virtual marshalling trains during emergency braking, the minimum safety distance of the trains should be as follows: ; The above calculations take into account the speed, acceleration, braking performance and relative position with the preceding train to ensure that under various operating conditions, a safe separation is maintained between the trains, avoiding collision risks.
10. The head space optimization control method for the follow-up operation of the virtual marshalling train according to claim 1 is characterized in that in step S5, specifically, speed errors between a pilot car and the follow-up car are monitored in real time through an algorithm, errors of a theoretical minimum safe head space and an actual operation head space and average values of the errors are calculated, and more accurate virtual marshalling train head space is obtained through comparison of speed error average values and variances under fuzzy PID and traditional PID algorithms and error average values and variances of the theoretical minimum safe head space and the actual operation head space.

Description

Head space optimal control method for virtual marshalling train following operation Technical Field The invention relates to the field of virtual marshalling operation of trains, in particular to an optimized control method for the locomotive spacing of the following operation of a virtual marshalling train. Background Along with the continuous increase of railway transportation density, especially in the high-frequency and high-density train running environment, the regulation and control of train intervals become key factors for improving transportation efficiency and guaranteeing driving safety. Most railway systems at present adopt a fixed marshalling train operation mode, and the problems of poor marshalling flexibility and inaccurate train interval adjustment exist. This approach is difficult to deal with complex and diverse operating environments, resulting in inefficient scheduling and possibly affecting the safe distance between trains. Virtual consist train scheduling can theoretically effectively solve this problem. By flexibly adjusting the train spacing according to the information of the real-time position, speed, acceleration and the like of the trains, the running efficiency of the trains can be optimized and the collision risk between the trains can be reduced. However, the implementation of the virtual marshalling train interval control algorithm in the prior art is not yet obvious, and particularly, how to control the train interval efficiently and accurately in a high-density train operation scene is still a difficult problem to be solved. Disclosure of Invention The invention aims to provide an optimal control method for the distance between the locomotive heads of a virtual marshalling train in following operation, which solves the problems. The technical scheme of the invention is that the locomotive spacing optimization control method for the following operation of the virtual marshalling train comprises the following steps: S1, acquiring and processing state data of a pilot vehicle, namely acquiring the state data of the position and the speed of the pilot vehicle and the line environment data of the line gradient and the curvature, synchronizing the state data and the line environment data to form a standardized data set, and providing basic input for the construction and the control execution of a follow-up algorithm; S2, constructing a variable domain fuzzy PID control algorithm based on virtual marshalling train following, namely based on the pilot vehicle state data and the line environment data output by S1, taking a train speed error and a position error of the pilot vehicle as input variables, inputting the input variables into a calculation module, constructing a variable domain fuzzy rule by the calculation module according to the input parameters, setting a telescopic factor on the input domain and a telescopic factor on the output domain, and calculating control parameters ,,; S3, designing an operation scene of the virtual marshalling train, namely independently configuring operation environment parameters of the virtual marshalling train as a preset operation environment of a subsequent control flow, wherein the operation environment parameters comprise an initial distance between a pilot train and a following train, a target operation speed interval, a line type and a marshalling state interval, and the line type comprises a straight section, a curve and a ramp; s4, speed and displacement processing of the following operation of the virtual marshalling train algorithm, wherein the speed and displacement processing is based on the real-time state data of S1 and the control parameters output by S2 ,,Calculating the target speed of the following vehicle through an algorithm, and determining the minimum vehicle head distance by combining a safety braking model to realize dynamic following control; S5, optimizing performance evaluation, namely based on the difference value of the speed output by S4 and the target speed and the difference value data of the following distance and the target distance, comparing the control precision and stability of the variable domain fuzzy PID and the traditional algorithm, and outputting performance optimization indexes; And S6, judging whether the process is finished according to the time step, namely judging whether the process reaches a preset operation time or a performance standard threshold by taking the sensor sampling step of the S1 as a basic period and combining the performance evaluation result of the S5, returning to the S1 to acquire the latest data again if the process is not standard, repeatedly executing the control flow of the S2-S5, and stopping the control and outputting a final optimization result if the process is standard. Preferably, in step S1, the method specifically includes: and acquiring dynamic state data of the pilot vehicle through various sensors, and transmitting the dynamic turntable data to a calculation