Automatic Return Method for a Water-Deficient Intelligent Driving Cleaning Vehicle

Abstract

The present invention relates to a method for automatically returning a water-deficient intelligent driving cleaning vehicle. The method comprises the following steps: Step 1: According to the ratio of the current weight of the water tank to the weight after the water tank is filled with water being less than 10%, a water-deficient signal is generated by a weight sensor at the lower part of the water tank and sent to an HMI controller through a CAN network; Step 2: The HMI controller performs signal recognition and sends it to an automatic driving controller through the CAN network. The present invention can not only realize the fully unmanned operation of an automatic driving sprinkler truck and can save manpower to a certain extent, but also has strong expected safety for automatic driving to return, and it is easy for manual intervention to cause safety accidents due to differences in operation levels.

Description

Technical Field

[0001] The present invention relates to the technical field of cleaning equipment, and particularly to a method for automatically returning a water-deficient intelligent driving cleaning vehicle. Background Art

[0002] Currently, autonomous driving sanitation vehicles have been demonstrated in multiple application scenarios, and among them, the intelligent sprinkling scenario is one of the several typical scenarios with the widest application range.

[0003] Currently, autonomous driving sprinkler trucks still adopt the man-machine co-driving mode for demonstration operations. Therefore, the current models on the market still rely on manual operation to judge whether to refill water by monitoring the water tank capacity on the instrument panel of the sprinkler truck. Its operation mode is that the autonomous driving sprinkler truck conducts route planning and automatic spraying operations, and the safety officer monitors the driving environment of the vehicle and the spraying water tank capacity. When the water tank capacity reaches the critical value, the safety officer suspends the vehicle's progress and spraying operations, and the safety officer operates the vehicle to return manually or automatically. The vehicle returns to the water injection point for water injection, and after the water injection is completed, the spraying operation continues. However, with the development of autonomous driving, the manual driving mode will ultimately be completely replaced by fully autonomous driving. In order to cope with the water volume monitoring process of the automatic sprinkler truck in this situation, an intelligent water injection monitoring system needs to be adopted. In the prior art, manual intervention is required for operation to return and inject water. With the development of autonomous driving technology, the driving safety officer will also no longer exist. Summary of the Invention

[0004] In view of the above deficiencies of the prior art, the present invention provides a method for automatically returning a water-deficient intelligent driving cleaning vehicle, which can not only realize the fully unmanned operation of the autonomous driving sprinkler truck and can save manpower to a certain extent, but also has strong expected safety for the autonomous driving return journey. Manual intervention is prone to safety accidents caused by differences in operation levels.

[0005] In order to achieve the above object and other related objects, the technical solution provided by the present invention is as follows:

[0006] A method for automatically returning a water-deficient intelligent driving cleaning vehicle includes the following steps:

[0007] Step 1: When the ratio of the current weight of the water tank to the weight of the water tank when it is full of water is less than 10%, a water shortage signal is generated by a weight sensor at the lower part of the water tank and sent to the HMI controller through the CAN network;

[0008] Step 2: The HMI controller performs signal recognition and sends it to the autonomous driving controller through the CAN network;

[0009] Step 3: After receiving the information, the autonomous driving controller determines whether the vehicle state meets the requirements for returning by the autonomous driving system. If not, the vehicle continues to drive and continuously sends the return signal to the autonomous driving controller until the autonomous driving controller determines that the requirements for returning are met;

[0010] Step 4: When the vehicle state meets the requirements for returning, the autonomous driving system issues a return command and publishes the execution layer command through the chassis VCU, and performs path planning based on the PRM algorithm and the A* algorithm to return to the water injection point.

[0011] Preferably, in Step 4, the PRM algorithm includes the following steps:

[0012] 1) Initialize, let G(V, E) be an undirected graph, where the vertex set V represents the collision-free state points of the vehicle, and the connection set E represents the collision-free connection path;

[0013] 2) Sample state points, sample collision-free state points in the vehicle driving path space and add them to the collision-free state points V;

[0014] 3) Neighborhood calculation, define the distance p. For the points that already exist in the collision-free state points V, if its distance from the collision-free points is less than p, then it is called the neighborhood point of the collision-free state point;

[0015] 4) Edge connection, connect the collision-free state points with their neighborhoods to generate connections;

[0016] 5) Collision detection, detect whether the connection collides with obstacles. If there is no collision, add it to the connection set E;

[0017] 6) End when all sampled points have completed the above steps, otherwise repeat 2-5.

[0018] Preferably, based on the PRM algorithm, multiple driving paths of the vehicle are obtained, and then combined with the A* algorithm to find the optimal path for the vehicle to drive.

[0019] Preferably, in Step 3, the vehicle state includes position, speed, temperature, equipment state, environment and route information.

[0020] Preferably, construct the state function of the vehicle:

[0021]

[0022] where, θ t is the model parameter at time t, J(θ) is the loss function, αt is the weight at time t, x 1 is the vehicle position parameter, x 2 is the vehicle speed parameter, x 3is the vehicle temperature parameter, x 4 is the vehicle equipment status parameter, x 5 is the vehicle route information parameter, x 6 is the vehicle environment parameter.

[0023] Preferably, based on the state function of the vehicle, a global optimal solution is obtained to determine whether the vehicle meets the requirements for returning.

[0024] Preferably, based on the ratio of the current weight of the water tank to the weight after the water tank is filled with water being less than 10%, and three or four or five or six of the vehicle state parameters conforming to the global optimal solution Bu, the vehicle returns to the water injection point according to the path planning.

[0025] The present invention has the following positive effects:

[0026] 1) The present invention can realize the fully unmanned operation of the automatic driving sprinkler truck.

[0027] 2) The present invention can save manpower to a certain extent.

[0028] 3) The expected safety of the automatic driving return of the present invention is strong, and manual intervention is prone to safety accidents caused by differences in operation levels. Description of the Drawings

[0029] Figure 1 is a schematic diagram of the working process of the present invention. Detailed Embodiments

[0030] The following will describe in detail the specific embodiments of the present invention with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for the purpose of illustration and explanation of the present invention, and are not intended to limit the present invention.

[0031] Embodiment: As Figure 1 shown, a method for automatically returning a water-deficient intelligent driving cleaning vehicle includes the following steps:

[0032] Step 1: Based on the ratio of the current weight of the water tank to the weight after the water tank is filled with water being less than 10%, the weight sensor at the lower part of the water tank generates a water shortage signal and sends it to the HMI controller through the CAN network;

[0033] Step 2: The HMI controller performs signal recognition and sends it to the automatic driving controller through the CAN network;

[0034] Step 3: After receiving the information, the automatic driving controller determines whether the vehicle state meets the requirements for returning through the automatic driving system. If not, the vehicle continues to drive and continuously sends a return signal to the automatic driving controller until the automatic driving controller determines that the requirements for returning are met;

[0035] Step 4: When the vehicle status meets the requirements for returning, the autonomous driving system issues a return command and releases an execution layer command through the chassis VCU, performs path planning based on the PRM algorithm and the A* algorithm, and returns to the water injection point.

[0036] Specifically, in Step 4, the PRM algorithm includes the following steps:

[0037] 1) Initialize by setting G(V, E) as an undirected graph, where the vertex set V represents the collision-free state points of the vehicle, and the connection set E represents the collision-free connection paths.

[0038] 2) Sample state points by sampling collision-free state points in the vehicle driving path space and adding them to the collision-free state point set V.

[0039] 3) Calculate neighborhoods. Define the distance p. For the points that already exist in the collision-free state point set V, if its distance from a collision-free point is less than p, then it is called a neighborhood point of the collision-free state point.

[0040] 4) Connect the edges by connecting the collision-free state points to their neighborhoods to generate connections.

[0041] 5) Perform collision detection by detecting whether the connections collide with obstacles. If there is no collision, add them to the connection set E.

[0042] 6) End when all sampled points have completed the above steps; otherwise, repeat steps 2 - 5.

[0043] Furthermore, based on the PRM algorithm, multiple driving paths of the vehicle are obtained, and then combined with the A* algorithm to find the optimal path for the vehicle to drive. The A* algorithm is the most effective direct search method for solving the shortest path in a static road network.

[0044] Specifically, the A* algorithm includes the following steps:

[0045] F1: Obtain the map information between the current position of the vehicle and the position of the water injection point, with the current position as the starting point and the water injection point as the ending point.

[0046] F2: Based on the map information, calculate the distances from the starting point to the surrounding points respectively, and output the point M corresponding to the shortest distance.

[0047] F3: Then, taking point M as the starting point, calculate the distances from point M to the surrounding points respectively, and output the point N corresponding to the shortest distance.

[0048] F4: Repeat steps F2 and F3 until the shortest path between the starting point and the ending point is found.

[0049] Specifically, the vehicle status includes position, speed, temperature, equipment status, environment, and route information, and a state function of the vehicle is constructed:

[0050]

[0051] where θ t is the model parameter at time t, J(θ) is the loss function, αt is the weight at time t, and x 1 is the vehicle position parameter, x 2 is the vehicle speed parameter, x 3 is the vehicle temperature parameter, x 4 is the vehicle equipment status parameter, x 5 is the vehicle route information parameter, x 6 is the vehicle environment parameter.

[0052] Specifically, based on the state function of the vehicle, a global optimal solution Bu is obtained, so as to determine whether the vehicle meets the requirements for returning.

[0053] Specifically, based on the ratio of the current weight of the water tank to the weight after the water tank is filled with water being less than 10%, and three or four or five or six of the vehicle state parameters conforming to the global optimal solution Bu, the vehicle returns to the water injection point according to the path planning.

[0054] Specifically, when the ratio of the current weight of the vehicle water tank to the weight after the water tank is filled with water is less than 10%, the vehicle state parameters include the position parameter, speed parameter, temperature parameter, equipment status parameter, environment parameter, and route information parameter. As long as more than three vehicle state parameters are satisfied, the vehicle's automatic driving system can determine that the vehicle can return, so the vehicle returns along the planned route for refueling.

[0055] In summary, the present invention can not only realize the completely unmanned operation of the automatic driving sprinkler truck and can save manpower to a certain extent, but also has strong expected safety for the automatic driving return journey. Manual intervention is prone to safety accidents caused by differences in operation levels.

[0056] In addition, any combination can be made between various different embodiments of the present invention, as long as it does not violate the idea of the embodiments of the present invention, it should also be regarded as the content disclosed by the embodiments of the present invention.

Claims

1. An automatic return method for a water - shortage intelligent driving cleaning vehicle, characterized in that, it includes the following steps: Step 1: When the ratio of the current weight of the water tank to the weight after the water tank is full of water is less than 10%, the weight sensor at the lower part of the water tank generates a water - shortage signal and sends it to the HMI controller through the CAN network; Step 2: The HMI controller performs signal recognition and sends it to the autonomous driving controller through the CAN network; Step 3: After receiving the information, the autonomous driving controller judges whether the vehicle state meets the return requirements through the autonomous driving system. If not, the vehicle continues to drive and continuously sends a return signal to the autonomous driving controller until the autonomous driving controller determines that the return requirements are met; Step 4: When the vehicle state meets the return requirements, the autonomous driving system issues a return command and releases an execution - layer command through the chassis VCU, and performs path planning based on the PRM algorithm and the A* algorithm to return to the water - filling point; In Step 4, the PRM algorithm includes the following steps: 1) Initialize, let G(V, E) be an undirected graph, where the vertex set V represents the collision - free state points of the vehicle, and the connection set E represents the collision - free connection path; 2) State - point sampling, sample collision - free state points in the vehicle driving path space and add them to the collision - free state points V; 3) Neighborhood calculation, define the distance p. For the points that already exist in the collision - free state points V, if its distance from the collision - free points is less than p, then it is called a neighborhood point of the collision - free state point; 4) Edge connection, connect the collision - free state points with their neighborhoods to generate connections; 5) Collision detection, detect whether the connection collides with an obstacle. If there is no collision, add it to the connection set E; end when all sampled points have completed the above steps, otherwise repeat 2 - 5; Based on the PRM algorithm, multiple driving paths of the vehicle are obtained, and then combined with the A* algorithm to find the optimal path for the vehicle to drive.

2. The automatic return method for a water - shortage intelligent driving cleaning vehicle according to claim 1, characterized in that, in Step 3, the vehicle state includes position, speed, temperature, equipment state, environment and route information.

3. The automatic return method for a water - shortage intelligent driving cleaning vehicle according to claim 2, characterized in that, construct a state function of the vehicle: ， where, θ t is the model parameter at time t, J(θ) is the loss function, α t is the weight at time t, x 1 is the vehicle position parameter, x 2 is the vehicle speed parameter, x 3 is the vehicle temperature parameter, x 4 is the vehicle equipment status parameter, x 5 is the vehicle route information parameter, x 6 is the vehicle environment parameter.

4. The automatic return method for a water - shortage intelligent driving cleaning vehicle according to claim 3, characterized in that: Based on the vehicle's state function, a global optimal solution B is obtained u , so as to determine whether the vehicle meets the requirements for returning 5. The automatic return method for a water - shortage intelligent driving cleaning vehicle according to claim 4, characterized in that: Based on the ratio of the current weight of the water tank to the weight after the water tank is filled with water being less than 10%, and the vehicle state parameters conforming to the global optimal solution B u Among three or four or five or six state parameters, the vehicle returns to the water injection point according to the path planning.

Images