An ecological driving method considering signal control uncertainty

Abstract

The application provides an ecological driving method considering signal control uncertainty, and belongs to the technical field of ecological driving, and comprises the following steps: 1) establishing a vehicle decision control model; 2) obtaining the probability that the traffic signal is green in a control time domain; 3) calculating a feasible time period; and 4) solving the vehicle decision control model. Compared with the prior art, the application is based on the chance constraint model predictive control theory, integrates the uncertainty of the traffic signal into the vehicle decision control model, expands the application range of the ecological driving technology, promotes the landing application of the ecological driving technology, and effectively reduces the environmental pollution and energy consumption of the traffic system.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of ecological driving, and particularly relates to an ecological driving method considering signal control uncertainty. BACKGROUND

[0002] Ecological driving is a driving mode aiming at energy saving and emission reduction, reducing fuel consumption and improving driving safety. By optimizing driving behavior and vehicle energy management, the negative impact on the environment is reduced. At the intersection, vehicles can communicate with the intersection signal through the Internet of Vehicles technology, obtain signal phase and timing data (SPaT, Signal Phase and Timing), realize accurate speed guidance, ensure passing during the green light, and reduce unnecessary idling waiting. These strategies not only help to improve fuel efficiency, but also alleviate intersection congestion and promote the overall sustainable development of road traffic.

[0003] Signal control uncertainty generally refers to the uncertainty of traffic signal control, which is specifically manifested in that the opening time and duration of the green light phase are uncertain. At the inductive traffic signal controlled intersection, signal control uncertainty is particularly obvious. The essence of inductive traffic signal control is a control strategy based on real-time traffic flow to dynamically adjust signal timing. Through the inductive coil or other detection devices installed on the road surface, the arrival of vehicles is monitored. When a vehicle is detected, the signal control system will combine the current traffic conditions to determine whether to give priority to the vehicles in that direction. Compared with the timing traffic signal control, the inductive traffic signal control has greater flexibility and responsiveness, which can significantly optimize the traffic efficiency of the intersection.

[0004] However, the existing ecological driving technology mostly assumes that the traffic signal is completely determined, and the complete signal phase and timing data (SPaT, Signal Phase and Timing) can be obtained, so it is not suitable for inductive traffic signal controlled intersections, and thus cannot be implemented in large-scale practical applications, which seriously restricts the further energy saving and emission reduction of the transportation system. Therefore, the existing technology has defects and needs to be improved by targeted invention. SUMMARY

[0005] The purpose of the present application is to provide an ecological driving method considering signal control uncertainty, as shown in Figure 1 characterized in that it comprises the following steps:

[0006] S1) establishing a vehicle decision control model based on model predictive control theory;

[0007] S2) obtaining the probability of the traffic signal being green in the control time domain based on the existing technology;

[0008] S3) Calculate the feasible time period based on the probability of green light of traffic signal;

[0009] S4) Solve the vehicle decision control model.

[0010] Further, in S1, the mathematical formula of the vehicle decision control model is as follows:

[0011]

[0012] S.t.

[0013] ξ k+1 = Aξ k + Bu k

[0014]

[0015] u min ≤ u k ≤ u max

[0016] wherein,

[0017] System dynamics Cost function Constraints

[0018] wherein, k is the index of control step; K is the total number of control steps;

[0019] ukis the control quantity; ξ k is the state vector; is the expectation of the state vector;

[0020] is the matrix transposition operator;

[0021] Q is the weight matrix of the state vector; r is the weight of the control quantity; A is the state matrix; B is the control matrix;

[0022] is the minimum value of the state vector; is the maximum value of the state vector; u min is the minimum value of the control quantity; u max is the maximum value of the control quantity;

[0023] s k is the vehicle position; v k is the vehicle speed; a k is the acceleration;

[0024] △τ is the length of adjacent control steps; v d is the expected speed; q v is the speed weight;

[0025] is the minimum value of the vehicle position; is the maximum value of the vehicle position;

[0026] v max is the road speed limit;

[0027] -inf represents negative infinity; inf represents positive infinity; -inf and inf mean no constraint;

[0028] t0 is the initial control step, i.e. the initial time;

[0029] G(t0) represents the state of the traffic signal at t0; G(t0) = 1 means that the state of the traffic signal at t0 is green; G(t0) = 0 means that the state of the traffic signal at t0 is not green;

[0030] K pass is the feasible time period;

[0031] s stop is the position of the traffic light; s k and s stop are shown in Figure 2 ;

[0032] a min is the minimum value of the acceleration; a max is the maximum value of the acceleration.

[0033] Further, in S2, the control time domain refers to the control time domain T of the vehicle decision control model in S1, and the calculation formula is:

[0034] T = K△T

[0035] It is assumed that the traffic signal only performs one light color switching within T, as shown in Figure 3 ; the probability that the traffic signal is green at the kth step is , which is a set value or obtained by prior art.

[0036] Further, in S3, the feasible time period is calculated, specifically including the following steps:

[0037] S31) calculating the switching probability of the traffic signal at the kth step , as shown in Figure 4 , and the calculation formula

[0038] is as follows:

[0039]

[0040] wherein, P(k) represents the probability that the kth step is green; P(k) represents the probability that the kth step is red;

[0041] P(k) represents the probability that the kth step is red; P(k) represents the probability that the kth step is green.

[0042] S32) Calculate the switching probability F(k) of the traffic signal in the kth step, as follows:

[0043]

[0044] S33) Define the feasible time period K pass , mathematically expressed as follows:

[0045]

[0046] where k sw represents the upper or lower limit of K pass . It means that when G(t0) = 1, the control steps from the 0th step to the k sw th step correspond to the green light phase, and thus are feasible, while the control steps from the k sw th step to the K sw th step correspond to the red light phase, and thus are infeasible. Similarly, when G(t0) = 0, the control steps from the 0th step to the k sw th step correspond to the red light phase, and thus are infeasible, while the control steps from the k pass th step to the K pass th step correspond to the green light phase, and thus are feasible.

[0047] S34) User-defined risk coefficient γ, meaning as follows:

[0048] γ represents the confidence of the vehicle successfully passing through the intersection within K pass ; further, when G(t0) = 1, the formula expresses "the confidence of the vehicle successfully passing through the intersection within K pass at least γ" as F(k sw ) ≤ 1-γ;

[0049] when G(t0) = 0, the formula expresses "the confidence of the vehicle successfully passing through the intersection within K pass at least γ" as F(k sw ) ≥ γ.

[0050] S35) Calculate k sw , as follows:

[0051]

[0052] where F -1 is the inverse function of F.

[0053] Further, in S4, the open-source solver (such as OSQP) is called to solve the vehicle decision control model, and the specific calling method can be found in the official user manual, which is not described here.

[0054] Compared with the prior art, the present application has the following beneficial effects:

[0055] The innovation of the present application is that the vehicle decision control model considering signal control uncertainty is innovatively established based on the opportunity constraint model predictive control theory, thereby expanding the application range of the eco-driving technology. The superiority of the present application is that the eco-driving method provided by the present application is not only suitable for fixed timing traffic signal control intersections, but also suitable for signal control uncertain environments, such as inductive traffic signal control intersections, so that the present application effectively promotes the landing application of the eco-driving technology. The threshold of the present application is that the present application applies advanced V2X technology, and at the same time needs to deeply understand and flexibly apply mathematical theories such as model predictive control and opportunity constraint, so that the present application has a high threshold. BRIEF DESCRIPTION OF DRAWINGS

[0056] Figure 1 The technical roadmap of the eco-driving method considering signal control uncertainty of the present application.

[0057] Figure 2 The schematic diagram of s k and s stop .

[0058] Figure 3 The schematic diagram of light color switching.

[0059] Figure 4 The calculation schematic diagram of the switching probability of the traffic signal at the kth step . DETAILED DESCRIPTION

[0060] The eco-driving method considering signal control uncertainty of the present application will be described in more detail below with reference to the accompanying drawings, in which the preferred embodiments of the present application are represented, it should be understood that the person skilled in the art can modify the present application described herein while still achieving the advantageous effects of the present application, therefore, the following description should be understood as extensive knowledge for the person skilled in the art, and not as a limitation of the present application.

[0061] As shown in Figure 1 , the eco-driving method considering signal control uncertainty specifically includes the following steps:

[0062] Step 1

[0063] Based on the model predictive control theory, the vehicle decision control model is established, and the mathematical formula is represented as:

[0064]

[0065] S.t.

[0066] ξ k+1 = Aξ k + Bu k

[0067]

[0068] u min ≤ u k ≤ u max

[0069] where,

[0070] System dynamics Cost function Constraints

[0071] where, k is the index of control step; K is the total number of control steps;

[0072] u k is the control variable; ξ k is the state vector; is the expectation of the state vector;

[0073] is the matrix transpose operator;

[0074] Q is the weight matrix of the state vector; r is the weight of the control variable; A is the state matrix; B is the control matrix;

[0075] is the minimum value of the state vector; is the maximum value of the state vector; u min is the minimum value of the control variable; u max is the maximum value of the control variable;

[0076] s k is the vehicle position; v k is the vehicle speed; a k is the acceleration; △τ is the length of adjacent control steps; v d is the expected speed; q v is the speed weight;

[0077] is the minimum value of the vehicle position; is the maximum value of the vehicle position; v max is the road speed limit;

[0078] -inf represents negative infinity; inf represents positive infinity; -inf and inf mean no constraint;

[0079] t0 is the initial control step, i.e. the initial moment; G(t0) represents the state of the traffic signal at t0; G(t0) = 1 indicates that the state of the traffic signal at t0 is green;

[0080] G(t0) = 0 indicates that the state of the traffic signal at t0 is not green;

[0081] K pass is the feasible time period; s stop is the position of the traffic signal; s k and s stop are shown in the schematic diagram as Figure 2 a min is the minimum value of acceleration; a max is the maximum value of acceleration.

[0082] Step 2

[0083] Based on the prior art, the probability of the traffic signal being green within the control time domain is obtained.

[0084] The control time domain refers to the control time domain T of the vehicle decision control model in S1, and the calculation formula is:

[0085] T = K△τ

[0086] It is assumed that within the range of T, the traffic signal only performs one light color switching, as shown in Figure 3 The probability of the traffic signal being green at the kth step is a set value or obtained by prior art.

[0087] Step 3

[0088] Based on the probability of the traffic signal being green, the feasible time period is calculated. The specific steps are as follows:

[0089] (1) Calculate the switching probability of the traffic signal at the kth step The schematic diagram is shown in Figure 4 , and the calculation formula is as follows:

[0090]

[0091] wherein,

[0092]

[0093] P(k) represents the probability of being green at the kth step; P(k+1) represents the probability of being red at the kth step; P(k-1) represents the probability of being red at the kth step; P(k+1) represents the probability of being green at the kth step.

[0094] (2) Calculate the switching probability F(k) of the traffic signal in the k-th step. The calculation formula is as follows:

[0095]

[0096] (3) Define the feasible time period K pass The mathematical expression is as follows:

[0097]

[0098] Where, k sw K represents pass The upper or lower bound. This means that when G(t0) = 1, from step 0 to step k... sw The control step within the step corresponds to the green light phase, so it is feasible, and from the k-th step... sw The control step within the Kth step corresponds to the red light phase, therefore it is not feasible. Similarly, when G(t0) = 0, from the 0th step to the kth step... sw The control step within the k-th step corresponds to the red light phase, therefore it is not feasible. sw The control step within the Kth step corresponds to the green light phase, so it is feasible.

[0099] (4) User-defined risk coefficient γ, the meaning of which is as follows:

[0100] γ represents K pass The ability to ensure vehicles pass smoothly through the intersection; furthermore, when G(t0) = 1, the formula can be used to express "at K pass The probability that vehicles can successfully pass through the intersection is at least γ” for F(k) sw )≤1-γ; When G(t0)=0, express "in K pass The probability that vehicles can successfully pass through the intersection is at least γ” for F(k) sw )≥γ.

[0101] (5) Calculate k sw The calculation formula is as follows:

[0102]

[0103] Among them, F -1 It is the inverse function of F.

[0104] Step 4

[0105] Solve the vehicle decision control model.

[0106] You can call open-source solvers (such as OSQP) to solve the vehicle decision control model. For specific calling methods, please refer to the official user manual, which will not be elaborated here.

[0107] The above merely describes the preferred embodiments of the present application and does not limit the present application in any way. Any person skilled in the art can make any form of equivalent replacement or modification to the technical solutions and technical contents disclosed by the present application without departing from the scope of the technical solutions of the present application, and such changes still belong to the protection scope of the present application.

Claims

1. An eco-driving method considering signal control uncertainty, characterized by, The method comprises the following steps: S1) establishing a vehicle decision control model; S2) obtaining the probability of a green light in a control time domain; S3) calculating a feasible time period; S4) solving the vehicle decision control model; In S1, the mathematical formula of the vehicle decision control model is as follows: ； s.t. ； ； ； Wherein, ； ； ； wherein is an index of the control step; is the total number of control steps; is the control quantity; is the state vector; is the expectation of the state vector; is the matrix transpose operator; is a weight matrix of the state vector; is a weight of the control quantity; is a state matrix; is a control matrix; is the minimum value of the state vector; is the maximum value of the state vector; is the minimum value of the control; is the maximum value of the control; is a vehicle position; is a vehicle speed; is an acceleration; is the length of the adjacent control step; is the desired speed; is the speed weight; is a minimum value of the vehicle position; is a maximum value of the vehicle position; is the road speed limit; represents negative infinity; represents positive infinity; and means no constraint; is the initial control step, i.e. the initial time instant; indicates the state of the traffic signal at the time instant; indicates that the state of the traffic signal is green at the time indicates that the state of the traffic signal is green at the time indicates that the state of the traffic signal is not green at the time indicates that the state of the traffic signal is not green at the time is a feasible time period; is the position of a traffic light; is a minimum value of the acceleration; is a maximum value of the acceleration.

2. The eco-driving method of claim 1, wherein, In the S2, the control time domain is the control time domain of the vehicle decision control model in the S1 The calculation formula is: ； Assume that the traffic signal is green at and only one light color switching is performed, the probability that the traffic signal is green at the first step is obtained and denoted as .

3. The eco-driving method of claim 2, wherein, In S3, the calculation of the feasible time period specifically comprises the following steps: S31) calculating the first the probability of switching of the traffic signal at the step , the calculation formula is as follows: ； wherein ; ； represents the probability that the first step is green; represents the probability that the first step is red; represents the probability that the first the probability that the red light is on in step i; represents the probability that the first the probability that the green light is on in step i; S32) calculating the first Switching probability of the traffic signal in the step , the calculation formula is as follows: ； S33) defining a feasible time period which is mathematically expressed as follows: ； wherein represents the upper or lower limit; S34) User-defined risk factor ; in, Indicates in Ensuring vehicles can smoothly pass through intersections; and then, when When, express "in" using a formula The chances of vehicles successfully passing through the intersection are at least [missing information]. "for ;when When, express "in" using a formula The chances of vehicles successfully passing through the intersection are at least [missing information]. "for ; S35) Calculate The calculation formula is as follows: ； wherein is the inverse function of

Images