Vehicle accelerator brake control method and device based on mixed integer programming

Abstract

The invention provides a vehicle accelerator braking control method and device based on mixed integer programming, and relates to the technical field of vehicle intelligent driving. The vehicle accelerator braking control method based on mixed integer programming comprises the steps that according to driving state data of a target vehicle at the current moment and dynamic road information data and traffic situation sensing data of a preset road section in front of the target vehicle, a non-linear programming algorithm is utilized, obtaining a reference speed curve when the target vehicle runs on the preset road section; constructing an objective function and constraint conditions based on the reference speed curve; and based on the constraint condition, solving the target function to obtain the vehicle motion state parameters of the target vehicle at the plurality of look-ahead moments in the driving process of the target vehicle on the preset road section, so as to indicate the target vehicle to drive on the preset road section based on the vehicle motion state parameters of the plurality of look-ahead moments. According to the invention, intelligent control of the accelerator / brake of the vehicle can be safely and comfortably carried out with lower energy consumption.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent driving of vehicles, and in particular to a vehicle throttle brake control method and device based on mixed integer programming. BACKGROUND

[0002] In the automobile industry revolution, intelligentization is the core driving force, and vehicle intelligentization has more practical significance because it can improve operational efficiency, economic benefit and vehicle safety. The safety and comfort of intelligent driving speed control as a popular research topic in the automobile industry depend on whether the throttle / brake control system is smooth, stable and predictable. The traditional adaptive cruise control (ACC) system has obvious limitations, relying only on throttle control speed, which is only suitable for high-speed driving and is difficult to play a role in traffic congestion or urban driving scenarios. To solve this problem, researchers have developed an ACC extended system with Stop&Go function, which enables the vehicle to automatically control the throttle and brake pedal, and plays a significant role in urban driving and dense traffic conditions, especially in preventing and avoiding rear-end collisions and major accidents.

[0003] However, in the related art, vehicle throttle / brake control faces two major difficulties: first, the intelligent driving system needs to travel according to the given reference speed, and the appropriate reference speed curve is time-varying. Improper control strategy can cause throttle / brake shaking and switching, affecting driving sensation and safety, and most intelligent driving algorithms do not fully consider relevant factors. Second, the fuel saving capability of commercial truck intelligent driving system is a core evaluation index, requiring to reduce brake triggering, intensity and duration under the premise of ensuring safety and comfort, but most algorithms do not integrate the optimization goal of reducing brake energy loss when planning speed.

[0004] Therefore, how to safely and comfortably control the vehicle throttle / brake with low energy consumption is a technical problem to be solved. SUMMARY

[0005] The present application provides a vehicle throttle brake control method and device based on mixed integer programming to solve the above-mentioned defects in the prior art and achieve safe, comfortable and intelligent control of vehicle throttle / brake with low energy consumption.

[0006] The present application provides a vehicle throttle brake control method based on mixed integer programming, comprising the following steps.

[0007] According to the driving state data of the target vehicle at the current moment, and the dynamic road information data and traffic situation perception data of the preset road section in front of the target vehicle, a reference speed curve of the target vehicle when driving on the preset road section is obtained by using a nonlinear programming algorithm; a target function and a constraint condition are constructed based on the reference speed curve; the target function includes a penalty term for a vehicle motion state parameter deviating from a corresponding standard parameter, and a braking energy loss cost term; the standard parameter corresponding to the vehicle motion state parameter is determined according to the reference speed curve; the target function aims to minimize the sum of the penalty term and the braking energy loss cost term contained therein; based on the constraint condition, the target function is solved to obtain vehicle motion state parameters of the target vehicle at multiple lookahead times during driving on the preset road section, so as to instruct the target vehicle to drive on the preset road section based on the vehicle motion state parameters at the multiple lookahead times.

[0008] According to the vehicle throttle and brake control method based on mixed integer programming provided by the application, the vehicle motion state parameters include speed, acceleration and jerk; the standard parameters corresponding to the vehicle motion state parameters include standard speed, standard acceleration and standard jerk; the penalty term includes a first tracking error between speed and standard speed quantified by a quadratic cost function, a second tracking error between acceleration and standard acceleration quantified by a quadratic cost function, and a third tracking error between jerk and standard jerk quantified by a quadratic cost function.

[0009] According to the vehicle throttle and brake control method based on mixed integer programming provided by the application, the braking energy loss cost term is determined by using the following method: the braking energy loss cost term is determined according to the braking acceleration and speed of the target vehicle, and by using a positive definite matrix.

[0010] According to the vehicle throttle and brake control method based on mixed integer programming provided by the application, the jerk in the target function is expressed as a function of acceleration by using a difference operator; and the speed in the target function is expressed as a function of acceleration by using an accumulation operator, so as to convert the optimization variables of the target function into the acceleration of the target vehicle at the multiple lookahead times.

[0011] According to the present invention, a vehicle throttle-brake control method based on mixed integer programming is provided. The constraint conditions include acceleration constraints, which are: acceleration is not less than coasting acceleration and not greater than the upper limit of braking acceleration. The acceleration constraints are expressed as inequalities including acceleration, coasting acceleration, the upper limit of braking acceleration, and a sequence of throttle and brake indications, so as to introduce integer constraints into the acceleration constraints through the sequence of throttle and brake indications. The sequence of throttle and brake indications includes the following discrete variables: throttle indication, brake indication, and coasting indication.

[0012] According to the present invention, a vehicle throttle braking control method based on mixed integer programming is provided, wherein the acceleration constraint conditions are as follows: in, The throttle brake indication sequence, For acceleration, The gliding acceleration is... This is the upper limit of the braking acceleration.

[0013] According to the present invention, a vehicle throttle-brake control method based on mixed integer programming is provided. The method involves solving the objective function based on the constraints to obtain vehicle motion state parameters of the target vehicle at multiple look-ahead moments during its travel on a preset road segment. This includes: traversing various combinations of throttle-brake indicator signals at each look-ahead moment using a branch-and-bound method; for each combination of throttle-brake indicator signals, determining the cost value of the objective function under the acceleration constraints determined by the indicator combinations using a regularized smooth Fischer-Bermeister algorithm based on the reference speed curve; and determining the vehicle motion state parameters of the target vehicle at multiple look-ahead moments during its travel on the preset road segment using a regularized smooth Fischer-Bermeister algorithm based on the reference speed curve under the target acceleration constraints. The target acceleration constraints are determined by the combination of throttle-brake indicator signals with the minimum cost value.

[0014] According to the present invention, a vehicle throttle braking control method based on mixed integer programming is provided. The method involves traversing various combinations of throttle braking indicator sequences at each look-ahead time using a branch-and-bound approach, including: constructing a hierarchical decision tree; wherein the root node of the hierarchical decision tree represents a combination of indicator quantities without considering acceleration constraints, each sub-layer corresponds to a look-ahead time, and the values ​​of its sub-nodes represent a combination of throttle braking indicator sequences at each look-ahead time from the initial look-ahead time to the look-ahead time corresponding to the sub-layer; traversing the hierarchical decision tree in a depth-first manner; and for each combination of throttle braking indicator quantities in the sequence, using a regularized smoothed Fischer-Bermeister algorithm based on a reference speed curve to determine the cost value of the objective function under the acceleration constraints determined by the indicator combination, including: for each node of the traversed hierarchical decision tree corresponding to the indicator combination, using a regularized smoothed Fischer-Bermeister algorithm based on the reference speed curve to determine the cost value of the objective function under the acceleration constraints determined by the indicator combination.

[0015] According to the present invention, a vehicle throttle braking control method based on mixed integer programming is provided, wherein constructing a hierarchical decision tree includes: establishing the root node of the hierarchical decision tree; establishing child nodes layer by layer from the root node; wherein, pruning operations are performed on child nodes whose combination of indications changes within a preset time period.

[0016] According to the present invention, a vehicle throttle braking control method based on mixed integer programming is provided, wherein traversing the hierarchical decision tree in a depth-first manner includes: If the combination of indicators with the minimum value determined in the previous moment is located on the target branch of the hierarchical decision tree, the hierarchical decision tree is traversed starting from the branch at the same level and adjacent to the target branch; the target branch is the branch in the first level of the hierarchical decision tree where the node with the minimum value is located.

[0017] The present invention also provides a vehicle throttle braking control device based on mixed integer programming, comprising: The acquisition module is used to obtain a reference speed curve of the target vehicle traveling on the preset road segment based on the current driving status data of the target vehicle, dynamic road information data and traffic situation perception data of the preset road segment ahead of the target vehicle, using a nonlinear programming algorithm. The construction module is used to construct an objective function and constraints based on the reference speed curve. The objective function includes a penalty term for deviations of the vehicle motion state parameters from their corresponding standard parameters, and a braking energy loss cost term. The standard parameters corresponding to the vehicle motion state parameters are determined based on the reference speed curve. The objective function aims to minimize the sum of its penalty term and braking energy loss cost term. The solution module is used to solve the objective function based on the constraints to obtain the vehicle motion state parameters of the target vehicle at multiple look-ahead moments during its travel on the preset road segment, thereby instructing the target vehicle to travel on the preset road segment based on the vehicle motion state parameters at these multiple look-ahead moments.

[0018] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the vehicle throttle braking control method based on mixed integer programming as described above.

[0019] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the vehicle throttle braking control method based on mixed integer programming as described above.

[0020] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the vehicle throttle braking control method based on mixed integer programming as described above.

[0021] The present invention provides a vehicle throttle and brake control method and device based on mixed integer programming. By using nonlinear programming algorithms, a reference speed curve for the target vehicle while traveling on the preset road segment is obtained based on the current driving state data of the target vehicle, dynamic road information data of the preset road segment ahead of the target vehicle, and traffic situation perception data. Based on the reference speed curve, an objective function and constraints are constructed. Based on the constraints, the objective function is solved to obtain vehicle motion state parameters at multiple look-ahead moments during the target vehicle's travel on the preset road segment. This instructs the target vehicle to travel on the preset road segment based on the vehicle motion state parameters at multiple look-ahead moments. By considering the dynamic road information data and traffic situation perception data of the preset road segment ahead of the target vehicle, such as road slope and curvature, and whether there is a following vehicle, a reference speed curve is obtained. Based on this reference speed curve, a penalty term for deviations of the vehicle motion state parameters from their corresponding standard parameters and a braking energy loss cost term are solved. This results in vehicle motion state parameters that can suppress throttle / brake jitter, eliminate frequent throttle / brake switching, reduce fuel consumption, and improve driver comfort. This allows for safe, comfortable, and low-energy intelligent control of the vehicle's throttle and brakes. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating the vehicle throttle braking control method based on mixed integer programming provided by the present invention.

[0024] Figure 2 This is a flowchart illustrating the method for solving the objective function provided by the present invention.

[0025] Figure 3 This is an exemplary schematic diagram of the hierarchical decision tree provided by the present invention.

[0026] Figure 4 This is a schematic diagram of the vehicle throttle braking control device based on hybrid integer programming provided by the present invention.

[0027] Figure 5 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0029] The following is combined Figures 1-3 The present invention describes a vehicle throttle braking control method based on hybrid integer programming.

[0030] Figure 1 This is a flowchart illustrating the vehicle throttle braking control method based on mixed integer programming provided by the present invention, as shown below. Figure 1 As shown, the method includes the following: Step 101: Based on the current driving status data of the target vehicle, as well as the dynamic road information data and traffic situation perception data of the preset road section ahead of the target vehicle, a nonlinear programming algorithm is used to obtain the reference speed curve of the target vehicle when driving on the preset road section.

[0031] The target vehicle is the vehicle for which driving status analysis and speed planning are currently required. For example, in an intelligent driving scenario, the target vehicle is a commercial vehicle controlled or assisted by an intelligent driving system.

[0032] Driving status data refers to various data reflecting the current operating status of a target vehicle, including but not limited to information such as driving distance, speed, and acceleration. Driving distance indicates the distance the vehicle has traveled from its starting point to its current position; speed reflects the vehicle's current speed; and acceleration reflects the rate at which the vehicle's speed changes.

[0033] The dynamic road information data for the preset road segment ahead of the target vehicle refers to road-related data that changes over time within a specific distance (e.g., within 5 kilometers) ahead of the target vehicle. This data may include, but is not limited to: road gradient information, i.e., the degree of inclination of the road, which affects the resistance of the vehicle; road curvature, which describes the degree of curvature of the road and affects the vehicle's steering and driving stability; and speed limit information, which specifies the maximum speed that the vehicle is allowed to travel on this road segment.

[0034] Traffic situation awareness data is used to perceive the current traffic environment of a target vehicle. This data may include, but is not limited to, whether there is a vehicle ahead in the current lane, and how the presence of the vehicle ahead will affect the target vehicle's speed and following distance. By perceiving this traffic situation, speed planning can be made more in line with actual traffic conditions, thus avoiding rear-end collisions and other accidents.

[0035] The reference speed curve is a curve showing the speed that the target vehicle should follow over time when traveling on a preset road segment, planned using a nonlinear programming algorithm that combines the target vehicle's driving status data, dynamic road information data of the preset road segment ahead, and traffic situation perception data. The reference speed curve considers multiple factors and aims to minimize fuel consumption while meeting a given timeframe (such as arriving at the destination on time).

[0036] Step 102: Based on the reference velocity curve, construct the objective function and constraints.

[0037] The objective function includes a penalty term for deviations of the vehicle's motion state parameters from their corresponding standard parameters, and a braking energy loss cost term. The objective function aims to minimize the sum of its penalty term and braking energy loss cost term.

[0038] The standard parameters corresponding to the vehicle motion state parameters are determined based on the reference speed curve.

[0039] In some embodiments, vehicle motion state parameters include speed, acceleration, and jerk; the standard parameters corresponding to the vehicle motion state parameters include standard speed, standard acceleration, and standard jerk.

[0040] The reference speed curve itself provides the ideal speed values ​​that a vehicle should achieve at different times. These speed values ​​are optimized using a nonlinear programming algorithm based on factors such as the gradient and curvature of the road ahead, speed limits, and traffic conditions (e.g., whether there is traffic congestion ahead), aiming to minimize fuel consumption while meeting given time constraints. Therefore, each speed point on the reference speed curve represents a standard speed, and the vehicle should strive to maintain this standard speed during actual driving. For example, on a straight road section without obstruction from vehicles ahead, the reference speed curve might set a relatively high constant speed; this speed is the standard speed for the vehicle on that section of road.

[0041] Standard acceleration can be obtained by performing differential calculations on a reference velocity curve.

[0042] Standard jerk is obtained by further differential calculation of the standard acceleration parameters. Jerk reflects the rate of change of acceleration and has a significant impact on the ride comfort of the vehicle and the driver's perception.

[0043] In some embodiments, the penalty term included in the objective function comprises a first tracking error between the speed quantized using a quadratic cost function and the standard speed, a second tracking error between the acceleration quantized using a quadratic cost function and the standard acceleration, and a third tracking error between the jerk quantized using a quadratic cost function and the standard jerk. The braking energy loss cost term is determined based on the braking acceleration and speed of the target vehicle, using a positive definite matrix. The formula for the objective function is shown below: (1) in, , , , , is an N×N symmetric positive definite matrix. v , a , j These are the velocity, acceleration, and jerk to be optimized, respectively. This is the first tracking error; This is the second tracking error; The third tracking error helps to suppress throttle / brake vibration and frequent switching. This is a cost item for braking energy loss, used to reduce braking during driving and reduce overall fuel consumption; , , The errors are velocity error, acceleration error, and jerk error, respectively.

[0044] (2) (3) (4) in, For standard speed, For standard acceleration, For standard accelerometer.

[0045] Ψ is the braking acceleration, and its formula is as follows.

[0046] (5) in, The initial value of the vehicle's coasting acceleration at each look-ahead moment is provided by the system identification module. It is a small positive number used to ensure the smoothness of the formula.

[0047] Look-ahead time refers to selecting several points in time forward from the current moment during vehicle operation to predict and plan the vehicle's driving status over a future period. Look-ahead time is used to consider factors such as road information and traffic conditions ahead of the vehicle, allowing for more informed decisions regarding throttle and braking control, thus ensuring vehicle safety, comfort, and fuel economy. Assuming the current moment is t0, and N look-ahead time points are selected at 1-second intervals, then the look-ahead time can be represented as including: t 0、 t0+1, t0+2, ..., t0+N.

[0048] In some embodiments, the jerk in the objective function is expressed as a function of acceleration using the difference operator D; and the velocity in the objective function is expressed as a function of acceleration using the accumulation operator C, so as to transform the optimization variables of the objective function into the acceleration of the target vehicle at multiple look-ahead moments. This is expressed by the following formula.

[0049] (6) (7) The constraints include acceleration constraints, which state that the acceleration is not less than the coasting acceleration and not greater than the upper limit of the braking acceleration. This can be expressed by the following formula: (8) (9) (10) (11) Where s, v, and a represent the driving distance, speed, and acceleration, respectively. To determine the maximum driving distance allowed at the furthest forward look-in time. The maximum permissible speed is determined by taking into account the road speed limit, the curvature of the road ahead, and the following traffic situation. For gliding acceleration, This is the upper limit of braking acceleration.

[0050] Under acceleration constraints, the feasible region of decision variable 'a' is mostly separable, that is: (12) In some embodiments, the above acceleration constraints are expressed as inequalities including acceleration, coasting acceleration, upper bound of braking acceleration, and a sequence of throttle and brake indications, so as to introduce integer constraints into the acceleration constraints through the sequence of throttle and brake indications; wherein, the sequence of throttle and brake indications includes the following discrete variables: throttle indication, brake indication, and coasting indication. This is expressed by the following formula.

[0051] (13) in, This is a sequence of throttle and brake indicator values. For acceleration, For gliding acceleration, This is the upper limit of braking acceleration.

[0052] The throttle and brake indication sequence is a combination of discrete variables including throttle indication, brake indication, and coasting indication, used to describe the acceleration, braking, or coasting state that the vehicle should adopt at different look-ahead moments. The throttle and brake indication sequence is used to introduce integer constraints into the acceleration constraint conditions, assisting in constructing a mixed-integer programming problem to achieve throttle and brake control decisions. This can be expressed by the following formula: (14) in, The number of look-ahead moments, each look-ahead moment corresponds to a sequence of throttle and brake indications.

[0053] The brake indication value is -1 when the vehicle needs to brake in the forward-looking moment, and 0 otherwise. When the vehicle faces a situation requiring deceleration, such as encountering an obstacle ahead, a red light, or a downhill section, the brake indication value is -1, activating the braking system to slow the vehicle. This variable provides the intelligent driving system with clear braking commands, ensuring the vehicle can reduce speed promptly and safely. For example, when driving on a highway, if a vehicle ahead suddenly decelerates, the brake indication value changes to -1, and the vehicle brakes to avoid a rear-end collision. This indicates that the vehicle needs to brake.

[0054] The throttle indication value is 1 when the vehicle needs to accelerate in the forward-looking moment, and 0 otherwise. When the vehicle is in a situation requiring increased speed, such as starting from a standstill or overtaking at low speed, the throttle indication value is 1, triggering the throttle control system to increase engine power output and accelerate the vehicle. This variable translates the abstract need for vehicle acceleration into specific control commands. For example, This indicates that the vehicle needs to accelerate.

[0055] The coasting indicator, with a value of 0, indicates that the vehicle requires no additional power output and no braking, relying solely on inertia, when both the throttle and brake indicators are 0. When the vehicle is descending a slope at a safe speed, or slowly decelerating towards its destination, the coasting indicator helps the intelligent driving system maintain the vehicle's coasting state, reducing unnecessary throttle and braking operations and improving driving economy and smoothness. For example, on downhill sections of mountain roads, the vehicle can use the coasting indicator to enter a coasting state, reducing fuel consumption. This indicates that the vehicle needs to be in a coasting state.

[0056] Step 103: Based on the constraints, solve the objective function to obtain the vehicle motion state parameters at multiple look-ahead moments during the target vehicle's travel on the preset road segment, so as to instruct the target vehicle to travel on the preset road segment based on the vehicle motion state parameters at multiple look-ahead moments.

[0057] In the specific implementation process, the vehicle motion state parameters at multiple forward moments during the target vehicle's journey on the preset road segment can be determined based on the optimal solution of the objective function under acceleration constraints. These parameters include the target vehicle's speed, acceleration, and jerk at each forward moment.

[0058] For an example of solving the objective function based on constraints, see [link to example]. Figure 2 The relevant descriptions in the text will not be repeated here.

[0059] Figure 2 This is a flowchart illustrating the method for solving the objective function provided by the present invention, as shown below. Figure 2 As shown, the method includes the following: Step 201: Using the branch and bound method, traverse the various combinations of throttle and brake indicator signals in each look-ahead time.

[0060] Step 202: For each combination of throttle and brake indicator signals in the throttle and brake indicator sequence, the regularized smooth Fisher-Bermeister algorithm is used to determine the cost of the objective function under the acceleration constraint conditions determined by the indicator combination, based on the reference speed curve.

[0061] Branch and bound (BnB) is an optimization algorithm widely used in mixed integer programming problems. Its basic idea is to decompose the feasible region of the problem into multiple sub-regions (branches) and calculate the upper and lower bounds of the objective function for each sub-region. By comparing these upper and lower bounds, sub-regions that cannot contain the optimal solution can be eliminated (bounded), thereby gradually narrowing the search range and ultimately finding the global optimum. In the embodiments provided by this invention, branch and bound is used to efficiently traverse various reasonable combinations of throttle / brake instruction sequences to find the throttle and brake commands that satisfy the constraints and optimize the objective function.

[0062] The regularized and smoothed Fischer-Burmeister algorithm (FBRS) is an iterative algorithm for solving nonlinear optimization problems. It combines regularization and smoothing techniques of the Fischer-Burmeister function to improve stability and convergence speed. This algorithm is particularly suitable for handling optimization problems with equality and inequality constraints.

[0063] In the embodiments provided by the present invention, the FBRS algorithm is used to determine the cost of the objective function under the acceleration constraint conditions determined by the combination of indications based on the reference velocity curve.

[0064] For each combination of throttle and brake indication values ​​in the sequence, the acceleration constraints are determined. For example, when the throttle indication value is 1, the acceleration should be greater than the coasting acceleration; when the brake indication value is -1, the acceleration should be negative and less than the upper limit of the braking acceleration.

[0065] For each combination of indications and corresponding acceleration constraints, the FBRS algorithm is used to solve for the optimal value of the objective function.

[0066] The FBRS algorithm iteratively approximates the acceleration value that satisfies all constraints and minimizes the objective function. In each iteration, the algorithm updates the values ​​of the variables and recalculates the objective function until the convergence condition is met.

[0067] For each combination of indications, the FBRS algorithm outputs a cost value that reflects the combined effect of the deviation between the actual vehicle motion state and the reference motion state, as well as the braking energy loss, under that combination of indications. For example, the objective function shown in formula (1) has a cost value that reflects the magnitude of the first tracking error, the second tracking error, the third tracking error, and the braking energy loss cost term it includes.

[0068] By comparing the cost of different combinations of input parameters, the optimal sequence of throttle and brake commands can be selected to achieve smooth and stable vehicle operation and reduce fuel consumption.

[0069] In the specific implementation process, steps 201 to 202 are performed in the following manner.

[0070] Construct a hierarchical decision tree. The root node of the hierarchical decision tree contains a combination of indicators without considering acceleration constraints, i.e.: Each sub-layer corresponds to a look-ahead time, and the value of its lower node is a combination of throttle and brake indicator signals at each look-ahead time from the initial look-ahead time to the look-ahead time corresponding to the sub-layer.

[0071] For example only, such as Figure 3 The hierarchical decision tree shown has a root node value of , indicating a relaxed solution optimization that completely disregards acceleration constraints (i.e., neither acceleration nor braking). The root node is the starting point of the decision tree. From here, the hierarchical decision tree branches to different child nodes based on different combinations of throttle and brake inputs.

[0072] The hierarchical decision tree branches according to the possible states of the throttle and brake instruction sequence. Each sub-layer corresponds to a look-ahead time. Figure 3 The example shows the case where the maximum traversal depth is 5.

[0073] The number under each node represents the cost of the objective function corresponding to a combination of throttle / brake indicator sequences from the initial look-ahead time to the corresponding look-ahead time of that sub-layer. For example, in the left branch, node values ​​are 10, 12, 15, etc., which may represent the cost of the objective function under different combinations of indicator sequences.

[0074] Starting from the root node, the decision tree branches to the left or right based on the values ​​of the throttle and brake indicator sequence.

[0075] Left branch: Indicates a braking operation. For example, branching from root node 0 to node 1, then continuing to branch to node 3, then to node 4, and so on.

[0076] Right-hand branch: Indicates accelerated operation. For example, branching from root node 0 to node 2, then continuing to branch to node 4, then to node 8, and so on.

[0077] Figure 3 The lowest level node in the array is the leaf node, which represents the final state at the maximum traversal level.

[0078] The green node (value 7) is the node corresponding to the optimal solution of the objective function because it has the minimum cost while satisfying all acceleration constraints. The corresponding throttle and brake indication sequence takes the value [-1,-1,-1,1,1], representing the optimal combination of throttle and brake indications at each look-ahead time.

[0079] The hierarchical decision tree is traversed in a depth-first manner. For each node of the hierarchical decision tree, the combination of indices is used to determine the cost of the objective function under the acceleration constraint determined by the combination of indices, based on the reference velocity curve using the regularized smooth Fisher-Bermeister algorithm.

[0080] In the specific implementation process, a hierarchical decision tree is constructed in the following way: Establish the root node of the hierarchical decision tree; starting from the root node, establish child nodes layer by layer; among them, prune child nodes whose indicator combinations change within a preset time period. Since a reasonable accelerator / brake command sequence should not switch too frequently, when traversing the hierarchical decision tree, only child nodes whose interval between two consecutive switches is greater than the preset time period (e.g., 2 seconds) are traversed. This pruning operation can effectively reduce the search volume.

[0081] In some embodiments, if the combination of indicators with the minimum cost value determined in the previous moment is located on the target branch of the hierarchical decision tree, the hierarchical decision tree is traversed starting from the branch at the same level and adjacent to the target branch; the target branch is the branch in the first level of the hierarchical decision tree where the node with the minimum cost value is located.

[0082] In the embodiments provided by this invention, a search-based hot start is employed as described above, retaining the current throttle / brake indicator sequence in each algorithm cycle. When the next algorithm cycle begins, it first determines the Node corresponding to the previous throttle / brake indicator sequence. prev Is it on the branch corresponding to the one with the smaller cost value among the two nodes in the first level? If it is, then start from Node prev If you start from the beginning, traverse the entire tree in reverse order; otherwise, maintain the normal traversal order.

[0083] The embodiments provided by this invention can reduce the search workload by an average of 25% through search pruning and search hot-start. Furthermore, when performing FBRS optimization calculations for each node of the hierarchical decision tree, the main computational time is consumed in the operation of inverting the Hessian matrix. Because the problem has a high dimensionality, direct inversion is computationally expensive. The computational cost of the inversion operation can be effectively reduced by using the Woodbury formula and LLT Cholesky decomposition.

[0084] Step 203: Under the target acceleration constraint, the regularized smooth Fisher-Bermeister algorithm is used to determine the vehicle motion state parameters at multiple look-ahead moments during the target vehicle's journey on the preset road segment based on the reference velocity curve.

[0085] The target acceleration constraint is determined by the combination of throttle and brake indications that minimizes the cost value.

[0086] In the specific implementation process, by traversing the hierarchical decision tree, the combination of throttle and brake instruction sequences that minimizes the cost of the objective function is obtained. Based on the combination of instruction sequences, the target acceleration constraint is determined. Then, under this target acceleration constraint, the regularized smoothed Fisher-Bermeister algorithm is used to determine the vehicle motion state parameters at multiple look-ahead moments during the target vehicle's travel on the preset road segment, based on the reference speed curve. These parameters include the target vehicle's speed, acceleration, and jerk at each look-ahead moment.

[0087] In practical implementation, data from over 200 million kilometers of real-world driving across various vehicle models and road conditions demonstrates that the embodiments provided by this invention can efficiently generate throttle / brake commands that ensure driver safety and comfort in scenarios such as flat roads, mountain roads, tunnels, and following other vehicles. This effectively suppresses throttle / brake shudder and avoids frequent switching between throttle and brake. Furthermore, extensive real-world data shows that this method avoids unnecessary braking, only issuing braking commands when approaching speed limits or being close to the vehicle in front. This effectively limits braking energy loss and ensures fuel economy. In addition, the embodiments provided by this invention employ pruning, search-based hot-start, and optimized matrix calculation techniques to achieve low computational complexity, with computation time in the single-digit millisecond range on typical automotive chips, which effectively meets the real-time requirements of intelligent driving systems.

[0088] The vehicle throttle and brake control device based on mixed integer programming provided by the present invention will be described below. The vehicle throttle and brake control device based on mixed integer programming described below can be referred to in correspondence with the vehicle throttle and brake control method based on mixed integer programming described above.

[0089] Figure 4 This is a schematic diagram of the vehicle throttle braking control device based on hybrid integer programming provided by the present invention.

[0090] like Figure 4 As shown, the vehicle throttle-brake control device 400 based on mixed integer programming includes the following modules.

[0091] The acquisition module 410 is used to obtain the reference speed curve of the target vehicle when it is traveling on the preset road section based on the current driving status data of the target vehicle, the dynamic road information data and traffic situation perception data of the preset road section in front of the target vehicle, and the nonlinear programming algorithm.

[0092] The construction module 420 is used to construct an objective function and constraints based on the reference speed curve; wherein, the objective function includes a penalty term for deviation of the vehicle motion state parameters from their corresponding standard parameters, and a braking energy loss cost term; the standard parameters corresponding to the vehicle motion state parameters are determined according to the reference speed curve; the objective function aims to minimize the sum of its penalty term and braking energy loss cost term.

[0093] The solver module 430 is used to solve the objective function based on the constraints to obtain the vehicle motion state parameters of the target vehicle at multiple forward moments during its travel on the preset road segment, so as to instruct the target vehicle to travel on the preset road segment based on the vehicle motion state parameters at the multiple forward moments.

[0094] Figure 5 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 5 As shown, the electronic device may include: a processor 510, a communications interface 520, a memory 530, and a communications bus 540, wherein the processor 510, the communications interface 520, and the memory 530 communicate with each other through the communications bus 540. The processor 510 can call logic instructions in the memory 530 to execute a vehicle throttle braking control method based on mixed integer programming. This method includes: obtaining a reference speed curve for the target vehicle traveling on the preset road segment based on the current driving state data of the target vehicle, dynamic road information data and traffic situation perception data of the preset road segment ahead of the target vehicle, using a nonlinear programming algorithm; constructing an objective function and constraints based on the reference speed curve; wherein the objective function includes a penalty term for deviations of the vehicle motion state parameters from their corresponding standard parameters, and a braking energy loss cost term; the standard parameters corresponding to the vehicle motion state parameters are determined according to the reference speed curve; the objective function aims to minimize the sum of its penalty term and braking energy loss cost term; and solving the objective function based on the constraints to obtain vehicle motion state parameters at multiple forward time points during the target vehicle's travel on the preset road segment, thereby instructing the target vehicle to travel on the preset road segment based on the vehicle motion state parameters at the multiple forward time points.

[0095] Furthermore, the logical instructions in the aforementioned memory 530 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0096] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the vehicle throttle braking control method based on mixed integer programming provided by the above methods. The method includes: obtaining a reference speed curve of the target vehicle when driving on the preset road segment based on the current driving state data of the target vehicle, dynamic road information data and traffic situation perception data of the preset road segment ahead of the target vehicle, using a nonlinear programming algorithm; constructing an objective function and constraints based on the reference speed curve; wherein, the objective function includes a penalty term for deviation of the vehicle motion state parameters from their corresponding standard parameters, and a braking energy loss cost term; the standard parameters corresponding to the vehicle motion state parameters are determined according to the reference speed curve; the objective function aims to minimize the sum of its penalty term and braking energy loss cost term; and solving the objective function based on the constraints to obtain the vehicle motion state parameters of the target vehicle at multiple look-ahead moments during its driving on the preset road segment, so as to instruct the target vehicle to drive on the preset road segment based on the vehicle motion state parameters at the multiple look-ahead moments.

[0097] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements a vehicle throttle braking control method based on mixed integer programming provided by the above methods. The method includes: obtaining a reference speed curve of the target vehicle traveling on the preset road segment based on the current driving state data of the target vehicle, dynamic road information data and traffic situation perception data of the preset road segment ahead of the target vehicle, using a nonlinear programming algorithm; constructing an objective function and constraints based on the reference speed curve; wherein the objective function includes a penalty term for deviation of the vehicle motion state parameters from their corresponding standard parameters, and a braking energy loss cost term; the standard parameters corresponding to the vehicle motion state parameters are determined according to the reference speed curve; the objective function aims to minimize the sum of its penalty term and braking energy loss cost term; and solving the objective function based on the constraints to obtain the vehicle motion state parameters of the target vehicle at multiple look-ahead moments during its travel on the preset road segment, so as to instruct the target vehicle to travel on the preset road segment based on the vehicle motion state parameters at the multiple look-ahead moments.

[0098] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0099] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for vehicle acceleration and braking control based on mixed integer programming, characterized by, The method comprises the following steps: obtaining a reference speed curve of the target vehicle on a preset road section in front of the target vehicle according to driving state data of the target vehicle at a current time, dynamic road information data and traffic situation perception data of the preset road section; constructing a target function and a constraint condition based on the reference speed curve, wherein the target function comprises a penalty term for a vehicle motion state parameter deviating from a corresponding standard parameter and a braking energy loss cost term, and the standard parameter corresponding to the vehicle motion state parameter is determined according to the reference speed curve, and the target function aims to minimize the sum of the penalty term and the braking energy loss cost term contained therein; solving the target function based on the constraint condition to obtain vehicle motion state parameters of the target vehicle at multiple look-ahead times during driving on the preset road section, so as to instruct the target vehicle to drive on the preset road section based on the vehicle motion state parameters at the multiple look-ahead times.

2. The mixed integer programming based vehicle throttle brake control method of claim 1, wherein, The vehicle motion state parameters comprise speed, acceleration and jerk, and the standard parameters corresponding to the vehicle motion state parameters comprise standard speed, standard acceleration and standard jerk. The penalty term comprises a first tracking error between speed and standard speed quantified by a quadratic cost function, a second tracking error between acceleration and standard acceleration quantified by a quadratic cost function, and a third tracking error between jerk and standard jerk quantified by a quadratic cost function.

3. The mixed integer programming based vehicle throttle brake control method of claim 2, wherein, The braking energy loss cost term is determined in the following manner: The braking energy loss cost term is determined according to the braking acceleration and speed of the target vehicle and by using a positive definite matrix.

4. The mixed integer programming based vehicle throttle brake control method of claim 3, wherein, The jerk in the target function is expressed as a function of acceleration by using a difference operator; and The speed in the target function is expressed as a function of acceleration by using an accumulation operator, so as to convert the optimization variables of the target function into the acceleration of the target vehicle at the multiple look-ahead times.

5. The mixed integer programming based vehicle throttle brake control method of claim 4, wherein, The constraint condition comprises an acceleration constraint condition, and the acceleration constraint condition is: The acceleration is not less than the coasting acceleration and not greater than the upper limit of the braking acceleration. The acceleration constraint condition is expressed as an inequality containing acceleration, coasting acceleration, the upper limit of the braking acceleration and a throttle-brake indication sequence, so as to introduce an integer constraint into the acceleration constraint condition through the throttle-brake indication sequence, wherein the throttle-brake indication sequence contains the following discrete variables: throttle indication, brake indication and coasting indication.

6. The mixed integer programming based vehicle throttle brake control method of claim 5, wherein, The acceleration constraint condition is as follows: wherein, is the throttle brake command sequence, is the acceleration, is the coasting acceleration, is the brake acceleration upper bound.

7. The mixed integer programming based vehicle throttle brake control method of claim 6, wherein, The solving of the target function based on the constraint condition to obtain the vehicle motion state parameters of the target vehicle at the multiple look-ahead times during driving on the preset road section comprises: iterating various indication combinations of the throttle-brake indication sequence at each look-ahead time by using a branch and bound method. determining, for each of the throttle-brake indication combination of the throttle-brake indication sequence, a value of the objective function under an acceleration constraint determined by the indication combination according to the reference speed curve by using a regularized smooth Fisher-Burmeister algorithm; determining, according to the reference speed curve, vehicle motion state parameters of the target vehicle at multiple look-ahead time points during the driving on the preset road section by using a regularized smooth Fisher-Burmeister algorithm under a target acceleration constraint; wherein the target acceleration constraint is an acceleration constraint determined by the throttle-brake indication combination of the throttle-brake indication sequence with the minimum value.

8. The mixed integer programming based vehicle throttle brake control method of claim 7, wherein, The use of branch and bound method, traversing each look-ahead time throttle-brake indication sequence of various indication combinations, including: constructing a hierarchical decision tree; wherein the value of the root node of the hierarchical decision tree is the indication combination without considering the acceleration constraint, each sub-layer corresponds to a look-ahead time in turn, and the value of the node under the sub-layer is an indication combination of the throttle-brake indication sequence from the starting look-ahead time to the look-ahead time corresponding to the sub-layer, during which each look-ahead time is experienced; traversing the hierarchical decision tree in a depth-first manner; The determination, for each of the throttle-brake indication combination of the throttle-brake indication sequence, of the value of the objective function under the acceleration constraint determined by the indication combination according to the reference speed curve by using the regularized smooth Fisher-Burmeister algorithm, includes: for each node of the hierarchical decision tree corresponding to the indication combination, determining the value of the objective function under the acceleration constraint determined by the indication combination according to the reference speed curve by using the regularized smooth Fisher-Burmeister algorithm.

9. The mixed integer programming based vehicle throttle brake control method of claim 8, wherein, The construction of the hierarchical decision tree includes: establishing the root node of the hierarchical decision tree; starting from the root node, establish sub-nodes layer by layer; wherein, for the sub-nodes whose indication combinations change within a preset time length, pruning operation is performed.

10. The mixed integer programming based vehicle throttle brake control method according to any one of claims 8 to 9, characterized in that, The traversal of the hierarchical decision tree in a depth-first manner includes: if the indication combination with the minimum value determined at the last time point is located on the target branch of the hierarchical decision tree, start traversing the hierarchical decision tree from the branch adjacent to the target branch at the same level; the target branch is the branch corresponding to the node with the minimum value in the first layer of the hierarchical decision tree.

11. A vehicle throttle brake control device based on mixed integer programming, characterized by, including: an acquisition module, configured to obtain, according to driving state data of a target vehicle at a current time point and dynamic road information data and traffic situation perception data of a preset road section in front of the target vehicle, a reference speed curve of the target vehicle when driving on the preset road section by using a nonlinear programming algorithm; a construction module, configured to construct an objective function and constraint conditions based on the reference speed curve; wherein the objective function includes a penalty term for a vehicle motion state parameter deviating from a corresponding standard parameter and a brake energy loss cost term; the standard parameter corresponding to the vehicle motion state parameter is determined according to the reference speed curve; the objective function aims to minimize the sum of the penalty term and the brake energy loss cost term contained therein; A solving module is configured to solve the objective function based on the constraint condition, to obtain vehicle motion state parameters of the target vehicle at multiple look-ahead time points during driving of the target vehicle on the preset road section, so as to instruct the target vehicle to drive on the preset road section based on the vehicle motion state parameters at the multiple look-ahead time points.

12. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor implements the vehicle throttle brake control method based on mixed integer programming according to any one of claims 1 to 10 when executing the computer program.

13. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the vehicle throttle brake control method based on mixed integer programming according to any one of claims 1 to 10.

Images