A path planning method for an unmanned container transport vehicle

Abstract

The application discloses a path planning method for unmanned container transport vehicles, which comprises the following steps: an environment model is established by using a grid method according to an existing operation environment, an intuitive environment map is established, and an initial map is made; an initial action route of the vehicle is obtained by using an A* algorithm through the initial map; the initial action route is converted into specific action instructions of the vehicle, and the specific action instructions are converted into actual operation steps; a path planning algorithm of the improved A* is used to update the map and the path planning; the path planning algorithm of the improved A* improves a heuristic function, sets a function related to the number of obstacles near the vehicle, and adaptively changes according to the surrounding environment. Compared with the prior art, the heuristic function in the A* algorithm is optimized, the system working efficiency is improved, the quality of the generated path is higher, and the path planning function of the unmanned distribution vehicle is realized.

Description

Technical Field

[0001] This invention relates to the field of unmanned delivery equipment technology, and in particular to a path planning method for unmanned container transport vehicles. Background Technology

[0002] The control systems for driverless container trucks are currently undergoing continuous evolution and development. Utilizing artificial intelligence and automation technologies, these vehicles can drive and operate autonomously without human drivers, and are considered a crucial future direction for the logistics industry.

[0003] With the growth of international trade and the development of the logistics industry, the demand for more efficient, safe, and reliable transportation methods is constantly increasing. Driverless container transport vehicles can serve as a potential solution to these problems. They can reduce the risk of human error and accidents caused by human drivers, and improve the vehicle's transport capacity and efficiency. Furthermore, driverless vehicles can employ intelligent route planning and optimization to achieve energy conservation and environmental protection goals.

[0004] Path planning is crucial for autonomous container trucks. Traditional A* algorithms combine Dijkstra's algorithm and optimal optimization algorithms to ensure the optimal path is found, while also employing heuristic search to improve search efficiency. However, traditional A* algorithms don't consider the presence of various obstacles in the surrounding environment. In warehouse environments, A* algorithms suffer from slow node expansion, low search efficiency, and tortuous path generation. Therefore, it's essential to address how to implement heuristic functions to further improve the algorithm's planning efficiency based on the complex environment surrounding the vehicle. Summary of the Invention

[0005] Purpose of the invention: To address the problems existing in the prior art, this invention provides a path planning method for unmanned container transport vehicles. It optimizes the heuristic function in the A* algorithm, adaptively changes it according to the surrounding environment, improves the system's working efficiency, and generates higher quality paths.

[0006] Technical Solution: This invention proposes a path planning method for an unmanned container transport vehicle, comprising an unmanned vehicle body (hereinafter referred to as the vehicle), wherein ultrasonic sensors are installed around the unmanned vehicle body to sense the density of obstacles in the surrounding area; the path planning method includes the following steps:

[0007] Step 1: Based on the existing operating environment, an environment model is established using the grid method. The actual environment in which the car is located is abstracted and represented as a two-dimensional map. In the grid map, the car is regarded as a point mass moving in two-dimensional space. Black grids represent obstacle space, and white areas represent free space. Each grid corresponds to a unique two-dimensional coordinate, thereby establishing an intuitive environment map and making the initial map.

[0008] Step 2: Using the initial map obtained above, use the A* algorithm to obtain the initial route of the car;

[0009] Step 3: When the car obtains the initial route, it begins to move, converting the initial route into specific car action instructions. After receiving the action instructions, the car converts the specific action instructions into actual running steps.

[0010] Step 4: While the car moves along the initial route, the map and path planning are continuously updated using the improved A* path planning algorithm. The improved A* path planning algorithm assumes that the cost of any node can be calculated using a cost estimation function, and selects the node with the smallest cost value from among the nodes whose cost values ​​have been calculated as the next expansion point of the algorithm. If the target point is selected as the next expansion point, it means that the optimal path has been found. The cost estimation function of the improved A* path planning algorithm is:

[0011] F(n) = G(n) + Z(n) * H(n)

[0012] Where G(n) is the dissipation function, representing the actual cost from the starting node node_S to node n; H(n) is the heuristic function, representing the estimated cost from node n to the target node node_G; F(n) represents the estimated cost from the starting node to the target node via node n; and Z(n) is a function related to the number of obstacles near the vehicle.

[0013] Furthermore, the valuation function is as follows:

[0014] (1) If the heuristic function H(n) approaches 0, then the influence of H(n) on F(n) is ignored, then F(n) = G(n), and the A* algorithm becomes the Dijkstra algorithm. At this time, the algorithm can find the optimal path, but it needs to expand a large number of nodes, and the algorithm efficiency is low.

[0015] (2) If the dissipation function G(n) approaches 0, the influence of G(n) on F(n) is ignored. Then F(n) = H(n), and the A* algorithm becomes the GBFS algorithm. At this time, the algorithm cannot guarantee that it will find the optimal path, but it has fewer search nodes and higher search efficiency.

[0016] Furthermore, ultrasonic sensors mounted around the vehicle detect the density of obstacles around it. These sensors continuously detect obstacles and transmit the data to the vehicle. The number of obstacles detected by the sensors determines whether a threshold is exceeded. When the number of obstacles detected exceeds the threshold, it indicates a complex environment, making the optimal path to the target more important and requiring careful searching. Conversely, when the number of obstacles detected is below the threshold, it indicates a simple environment, prioritizing a faster arrival at the destination.

[0017] When the sensor detects obstacles near the car exceeding the threshold, it indicates that the car's surrounding environment is complex, and the optimal path to the target is more important. The value of Z(n) decreases, causing F(n) to tend towards Dijkstra's algorithm. When the sensor detects obstacles near the car below the threshold, it indicates that the car's surrounding environment is simple, and it is more important to reach the destination area quickly. The value of Z(n) increases, causing F(n) to tend towards GBFS algorithm.

[0018] Furthermore, when the surrounding environment of the car is complex and the optimal path to the target is more important, Z(n) becomes smaller, Z2(n)∈(0,1], x≥k / 2, Z2=1 / [x-(k / 2-1)];

[0019] When the surrounding environment of the car is simple and it is more important to reach the destination area quickly, the value of Z(n) increases, Z1(n)∈(1,10], x<k / 2, Z1=10-(18 / k)*x;

[0020] Where k is the number of directions detected by the sensor around the car, x is the number of directions with obstacles detected, and Z(n) is a function that reflects the number of obstacles near the car. Z(n) has two cases, namely Z2(n) and Z1(n). When the environment around the car is complex and there are many obstacles, exceeding the threshold, the best path to the target is more important, so Z(n) = Z2(n). When the environment around the car is simple and there are few obstacles, it is more important to reach the destination area quickly, so Z(n) = Z1(n).

[0021] Furthermore, the specific method for obtaining the initial movement route of the car in step 2 is as follows:

[0022] The initial state is to put the starting point into a closed set and all nodes directly connected to the starting point into an open set. Unless the ending point has already been put into a closed set, or the open set is empty (in which case there may be no feasible solution), repeat the following steps:

[0023] 1) Create a map representation: Divide the environment into discrete grids or nodes, and assign a state to each node;

[0024] 2) Initialize the start and end points: Determine the starting and target positions of the car and use them as input for the A* algorithm;

[0025] 3) Create open and closed lists: Open lists are used to store nodes to be explored, and closed lists are used to store nodes that have already been explored;

[0026] 4) Add the starting point to the open list: Add the starting point node to the open list as the starting point for exploration;

[0027] 5) Enter the main loop: Iterate through the open list until the endpoint node is found or the open list is empty;

[0028] 6) Find the current best node: Select the node with the lowest estimated total path cost from the open list as the current node;

[0029] 7) Explore neighboring nodes: For the neighboring nodes of the current node, calculate and update the actual path cost G from their origin to the current node and the estimated path cost H from the current node to the destination.

[0030] 8) Update nodes and paths: If an adjacent node is not in the open list, add it to the open list and set the current node as its parent node. If an adjacent node is already in the open list, update its G value and update the parent node according to the change in the G value.

[0031] 9) Determine if the destination has been reached: If the destination node appears in the open list, it means that the best path has been found and the algorithm ends.

[0032] Furthermore, step 3, which transforms the initial route into specific vehicle movement instructions, includes the following steps:

[0033] 1) Define the action set: Define the set of actions that the car can perform, including the basic actions of moving forward, backward, turning left, and turning right;

[0034] 2) Determine the relationships between nodes: Determine the relationships between nodes, that is, how the first node reaches the second node, how the second node reaches the third node, and so on;

[0035] 3) Calculate the angle and distance between nodes: For each pair of adjacent nodes, calculate the angle and distance the trolley needs to rotate and move based on the node's position information, which is achieved by calculating the Euclidean distance or straight-line distance between the nodes.

[0036] 4) Convert angles and distances into action commands: Convert angles and distances into actual action commands, convert angles into left or right turn angle values, and convert distances into forward or backward distance values ​​for the car.

[0037] Beneficial effects:

[0038] This invention employs an improved heuristic function method for the A* algorithm. The heuristic function in the A* algorithm is optimized and adaptively changed according to the surrounding environment, improving system efficiency and generating higher-quality paths. A function Z(n) related to the number of obstacles near the vehicle is set. The number of obstacles detected by ultrasonic sensors is used to determine whether a threshold is exceeded. When the number of obstacles detected by the sensors exceeds the threshold, it indicates that the surrounding environment of the vehicle is complex, and the optimal path to the target is more important, requiring careful searching for the optimal path. When the number of obstacles detected by the sensors is below the threshold, it indicates that the surrounding environment of the vehicle is simple, and the vehicle should quickly reach the destination area. Furthermore, the unmanned vehicle body can achieve free movement, obstacle detection, and target detection, and perform path planning based on the given target location, enabling the unmanned delivery vehicle to autonomously navigate and deliver goods to the pre-designated target location. Attached Figure Description

[0039] Figure 1 This is a flowchart of the method. Detailed Implementation

[0040] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0041] This invention discloses a path planning method for an unmanned container transport vehicle. It employs sensors, cameras, and radar to monitor the vehicle's surrounding environment in real time, avoiding collisions and accidents. Through precise positioning and path planning, it avoids congestion and route selection errors, reducing transportation time and resource waste. The method includes an unmanned vehicle body (hereinafter referred to as the vehicle), around which ultrasonic sensors are installed to detect the density of obstacles. Furthermore, the drive unit, actuators, and other supporting equipment on the vehicle are not the focus of this patent and will not be elaborated here. The method disclosed in this invention specifically includes the following steps:

[0042] Step 1: Establish an environment model using the grid method based on the existing operating environment. The primary task for global path planning of the robot is to establish an environment model to provide a search environment for path planning. Since the target is a small to medium-sized logistics warehouse, the operating environment is known. Therefore, the grid method is used for environment modeling, abstracting the actual environment in which the robot is located into a two-dimensional map, simplifying the operating space of the mobile robot. The actual environment in which the robot is located is abstracted into a two-dimensional map. In the grid map, the robot is regarded as a point mass moving in two-dimensional space. Black grids represent obstacle space, and white areas represent free space. Each grid corresponds to a unique two-dimensional coordinate, thus establishing an intuitive environment map and creating the initial map.

[0043] Step 2: Using the initial map obtained above, use the A* algorithm to obtain the initial route of the car.

[0044] The initial state of the A* algorithm is to put the starting point into a closed set and all nodes directly connected to the starting point into an open set. Unless the ending point has already been put into a closed set, or the open set is empty (in which case there may be no feasible solution), the following steps are repeated:

[0045] 1) Create a map representation: Divide the environment into discrete grids or nodes and assign a state to each node.

[0046] 2) Initialize the start and end points: Determine the starting and target positions of the car and use them as inputs for the A* algorithm.

[0047] 3) Create open and closed lists: Open lists are used to store nodes to be explored, and closed lists are used to store nodes that have already been explored.

[0048] 4) Add the starting point to the open list: Add the starting point node to the open list as the starting point for exploration;

[0049] 5) Enter the main loop: Iterate through the open list until the endpoint node is found or the open list is empty.

[0050] 6) Find the current best node: Select the node with the lowest estimated total path cost (F value) from the open list as the current node.

[0051] 7) Explore neighboring nodes: For the neighboring nodes of the current node, calculate and update the actual path cost G from their origin to the current node and the estimated path cost H from the current node to the destination.

[0052] 8) Update nodes and paths: If an adjacent node is not in the open list, add it to the open list and set the current node as its parent node. If an adjacent node is already in the open list, update its G value and update the parent node according to the change in the G value.

[0053] 9) Determine if the destination has been reached: If the destination node appears in the open list, it means that the best path has been found and the algorithm ends.

[0054] Step 3: When the car obtains the initial route, it begins to move, converting the initial route into specific car action instructions. After receiving the action instructions, the car converts the specific action instructions into actual running steps.

[0055] Converting each node on the path into actual action commands typically requires considering the vehicle's powertrain, controller, and navigation algorithm. The specific conversion process is as follows:

[0056] 1. Define the action set: First, you need to define the set of actions that the car can perform. For example, it can include basic actions such as forward, backward, left turn, and right turn.

[0057] 2. Determine the relationships between nodes: The path obtained from the A* algorithm is a series of nodes. We need to determine the relationships between the nodes, that is, how the first node reaches the second node, how the second node reaches the third node, and so on.

[0058] 3. Calculate the angles and distances between nodes: For each pair of adjacent nodes, calculate the angle the cart needs to rotate and the distance it needs to move based on the node's position information. This can be achieved by calculating the Euclidean distance or the straight-line distance between the nodes.

[0059] 4. Convert angles and distances into action commands: Based on the vehicle's power system and controller, convert angles and distances into actual action commands. Convert angles into left or right turn angle values, and convert distances into forward or backward distance values ​​for the vehicle.

[0060] After receiving an action command, the vehicle translates the specific action command into actual operating steps. Executing the action command: Based on the translated action command, the vehicle is controlled to perform the corresponding actions. This involves the vehicle's control system, including the motor drive and steering system.

[0061] Motor controllers or drivers are used to control the speed and direction of the motor. By sending control signals, parameters such as motor enable, rotation direction, and speed can be controlled, thereby enabling the car to move forward and backward. Feedback mechanisms and closed-loop control: To ensure the accuracy and stability of the motor drive, feedback devices such as encoders and position sensors can be used to detect the motor's position and speed in real time. Closed-loop control algorithms are then used to adjust the motor's movement to maintain the required speed and position.

[0062] The steering controller controls the angle of the steering mechanism. Based on steering commands, corresponding control signals are sent to adjust the steering mechanism's angle, enabling operations such as left turns, right turns, or straight-line movement. Precise turning maneuvers are achieved by controlling the steering mechanism's angle and speed, according to the vehicle's rotation radius and turning speed requirements. These are all standard techniques in the field of autonomous driving and will not be elaborated upon here.

[0063] Step 4: While the car is moving along the initial route, continuously update the map and path planning using the improved A* path planning algorithm.

[0064] As the vehicle moves along its initial route, the surrounding environment constantly changes due to the placement and transfer of goods. Therefore, the map and path planning need continuous updates to better adapt to the environment and improve the vehicle's transportation efficiency. Since the traditional A* algorithm does not consider the presence of various obstacles in the surrounding environment, an improved A* path planning algorithm is proposed. Ultrasonic sensors mounted around the vehicle detect the density of obstacles. These sensors continuously transmit information to the vehicle when they detect obstacles. The number of obstacles detected by the sensors determines whether a threshold is exceeded. When the number of obstacles detected exceeds the threshold, it indicates a complex environment, making the optimal path to the destination more important and requiring careful searching. Conversely, when the number of obstacles detected is below the threshold, it indicates a simple environment, prioritizing a faster arrival at the destination.

[0065] The improved A* algorithm is optimized as follows:

[0066] The traditional A* algorithm combines the ideas of Dijkstra's algorithm and optimal optimization algorithms. While ensuring that the optimal path can be found, it also uses heuristic search to improve the algorithm's search efficiency. The A* algorithm assumes that the cost of any node can be calculated using a cost estimation function, and selects the node with the smallest cost value from the nodes whose cost values ​​have been calculated as the next expansion point of the algorithm. If the target point is selected as the next expansion point, it means that the optimal path has been found.

[0067] F(n) and f are the estimated cost function (or valuation function) and function value of the current expanded node n, respectively; G(n) and g are the actual cost calculation function and function value from the starting node node_S to the current node n, respectively; H(n) and h are the estimated cost function (heuristic function) and function value from the current expanded node n to the target node node_G.

[0068] The improved A* path planning algorithm assumes that the cost of any node can be calculated using a cost estimation function. It then selects the node with the lowest cost among those with calculated costs as the next expansion point. If the target point is selected as the next expansion point, it indicates that the optimal path has been found. The traditional A* cost estimation function is:

[0069] F(n) = G(n) + H(n)

[0070] Where G(n) is the dissipation function, representing the actual cost from the starting node node_S to node n.

[0071] Where H(n) is the heuristic function, representing the estimated cost from node n to the target node node_G.

[0072] Where F(n) represents the estimated cost from the starting node to the target node via node n.

[0073] In the overall process of the A* algorithm, the core of the A* algorithm is the heuristic function, where G(n) is the cost from the starting point to the current point. Its value is generally fixed, so the optimization around the heuristic function is generally centered around H(n), that is, the prediction function.

[0074] The change of the heuristic function H(n) has a significant impact on the search speed and search results of the A* algorithm. Therefore, the choice of H(n) is a key aspect of the A* algorithm.

[0075] (1) If the heuristic function H(n) approaches 0, then the influence of H(n) on F(n) can be ignored, and F(n) = G(n), and the A* algorithm becomes the Dijkstra algorithm. At this time, the algorithm can always find the optimal path, but it needs to expand a large number of nodes, and the algorithm efficiency is low.

[0076] (2) If the dissipation function G(n) approaches 0, then the influence of G(n) on F(n) can be ignored, and F(n) = H(n), and the A* algorithm becomes the GBFS algorithm. At this time, the algorithm cannot guarantee that it will find the optimal path, but it searches fewer nodes and has high search efficiency.

[0077] The improved valuation function for A* is as follows:

[0078] F(n) = G(n) + Z(n) * H(n)

[0079] Z(n) is a function related to the number of obstacles near the car.

[0080] The improved heuristic function is actually an application of the dynamic weighting method. The principle is that at the beginning of the search, when the sensor detects obstacles near the car exceeding the threshold, it indicates that the car's surrounding environment is complex, and the best path to the target is more important. The value of Z(n) decreases, making F(n) tend towards Dijkstra's algorithm. When the sensor detects obstacles near the car below the threshold, it indicates that the car's surrounding environment is simple, and it is more important to reach the destination area quickly. The value of Z(n) increases, making F(n) tend towards GBFS algorithm.

[0081] When the surrounding environment of the car is complex and the best path to the target is more important, Z(n) becomes smaller, Z2(n)∈(0,1], x≥k / 2, Z2=1 / [x-(k / 2-1)].

[0082] When the surrounding environment of the car is simple and it is more important to reach the destination area quickly, the value of Z(n) increases, Z1(n)∈(1,10], x<k / 2, Z1=10-(18 / k)*x.

[0083] Where k is the number of directions detected by the sensor around the car, and x is the number of directions with obstacles detected. Z(n) is a function that reflects the number of obstacles near the car. Z(n) has two cases: Z2(n) and Z1(n). When the environment around the car is complex and there are many obstacles, exceeding the threshold, the best path to the target is more important, and Z(n) = Z2(n). When the environment around the car is simple and there are few obstacles, it is more important to reach the destination area quickly, and Z(n) = Z1(n).

[0084] The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They should not be construed as limiting the scope of protection of the present invention. All equivalent transformations or modifications made in accordance with the spirit and essence of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A path planning method for an unmanned container transport vehicle, comprising an unmanned vehicle body, hereinafter referred to as the vehicle, characterized in that: The autonomous vehicle is equipped with ultrasonic sensors around its perimeter to detect the density of obstacles in the surrounding area; the path planning method includes the following steps: Step 1: Based on the existing operating environment, an environment model is established using the grid method. The actual environment in which the car is located is abstracted and represented as a two-dimensional map. In the grid map, the car is regarded as a point mass moving in two-dimensional space. Black grids represent obstacle space, and white areas represent free space. Each grid corresponds to a unique two-dimensional coordinate, thereby establishing an intuitive environment map and obtaining the initial map. Step 2: Using the initial map, obtain the initial route of the car using the A* algorithm; Step 3: When the car obtains the initial route, it begins to move, converting the initial route into specific car action instructions. After receiving the action instructions, the car converts the specific action instructions into actual running steps. Step 4: While the car is moving along the initial route, update the map and path planning using the improved A* path planning algorithm; The improved A* path planning algorithm assumes that the cost of any node can be calculated using a cost estimation function. It then selects the node with the lowest cost among those with calculated costs as the next expansion point. If the target point is selected as the next expansion point, it indicates that the optimal path has been found. The cost estimation function of the improved A* path planning algorithm is: ； in, Let be the dissipation function, representing the dissipation function from the starting node. To the node The actual cost; For heuristic functions, represent nodes. To the target node The estimated cost; Indicates starting from the node and passing through the node The estimated cost to reach the target node. This is a function that reflects the number of obstacles near the car, where There are two situations, namely and ; When the surrounding ultrasonic sensors detect obstacles near the vehicle, they transmit the data to the vehicle. The number of ultrasonic waves detected by the sensors determines whether a threshold has been exceeded. When the number of obstacles detected by the sensors exceeds the threshold, it indicates that the surrounding environment of the vehicle is complex, and the optimal path to the target is more important, requiring careful searching for the optimal path. = , , , , The value decreases, making It tends to be similar to Dijkstra's algorithm; When the sensors detect that the number of obstacles near the vehicle does not exceed a threshold, it indicates that the vehicle's surrounding environment is simple, and quickly reaching the destination area is more important. = , , , , The value increases, making It tends towards the GBFS algorithm; in, To enable sensors to detect the number of directions around the car, The number of directions in which obstacles are detected.

2. The path planning method for an unmanned container transport vehicle according to claim 1, characterized in that: The cost estimation function is as follows: (1) If the heuristic function Approaching 0, at this point right If the effect is ignored, then The A* algorithm is transformed into Dijkstra's algorithm. At this point, the algorithm can find the optimal path, but it needs to expand a large number of nodes, resulting in low algorithm efficiency. (2) If the dissipation function Approaching 0, at this point right If the effect is ignored, then The A* algorithm is transformed into the GBFS algorithm. At this time, the algorithm cannot guarantee that it will find the optimal path, but it has fewer search nodes and higher search efficiency.

3. The path planning method for an unmanned container transport vehicle according to claim 1, characterized in that: The specific method for obtaining the initial movement route of the car in step 2 is as follows: The initial state is to put the starting point into a closed set and all nodes directly connected to the starting point into an open set. Unless the ending point has already been put into a closed set, or the open set is empty, repeat the following steps: 1) Create a map representation: Divide the environment into discrete grids or nodes, and assign a state to each node; 2) Initialize the start and end points: Determine the starting and target positions of the car and use them as input for the A* algorithm; 3) Create open and closed lists: Open lists are used to store nodes to be explored, and closed lists are used to store nodes that have already been explored; 4) Add the starting point to the open list: Add the starting point node to the open list as the starting point for exploration; 5) Enter the main loop: Iterate through the open list until the endpoint node is found or the open list is empty; 6) Find the current best node: Select the node with the lowest estimated total path cost from the open list as the current node; 7) Explore neighboring nodes: For the neighboring nodes of the current node, calculate and update the actual path cost G from their origin to the current node and the estimated path cost H from the current node to the destination. 8) Update nodes and paths: If an adjacent node is not in the open list, add it to the open list and set the current node as its parent node. If an adjacent node is already in the open list, update its G value and update the parent node according to the change in the G value. 9) Determine if the destination has been reached: If the destination node appears in the open list, it means that the best path has been found and the algorithm ends.

4. The path planning method for an unmanned container transport vehicle according to claim 1, characterized in that: Step 3, which converts the initial route into specific vehicle movement instructions, includes the following steps: 1) Define the action set: Define the set of actions that the car can perform, including the basic actions of moving forward, backward, turning left, and turning right; 2) Determine the relationships between nodes: Determine the relationships between nodes, that is, how the first node reaches the second node, how the second node reaches the third node, and so on; 3) Calculate the angle and distance between nodes: For each pair of adjacent nodes, calculate the angle and distance the trolley needs to rotate and move based on the node's position information, which is achieved by calculating the Euclidean distance between the nodes; 4) Convert angles and distances into action commands: Convert angles and distances into actual action commands, convert angles into left or right turn angle values, and convert distances into forward or backward distance values ​​for the car.

Images