Driving mode switching method and system based on different supervision modes and medium

Abstract

The invention relates to a driving mode switching method and system based on different supervision modes and a medium, and the method comprises the steps: U1, collecting the motion state data information of a vehicle under an in-vehicle supervision mode when the vehicle runs on a road, and collecting the motion state data information of the vehicle under a remote supervision mode; and U2, based on the motion state data information of the vehicle in the in-vehicle supervision mode, predicting the motion state of the vehicle by adopting a radial basis function neural network algorithm based on time sequence analysis, and obtaining the predicted motion state data information of the vehicle in the in-vehicle supervision mode. According to the invention, not only can the driving modes in different supervision modes be accurately switched, but also an inter-system safety redundancy mechanism is provided according to two different supervision modes of in-vehicle supervision and remote supervision, and the operation efficiency is ensured while the operation safety is ensured.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to a driving mode switching method, system and medium based on different supervision modes. Background Art

[0002] In the prior art, a Chinese patent (application number: 202110335072.1, publication number: CN 112947482A) discloses a method and system for hot switching between automatic driving and remote driving, the method comprising: when the driving mode is automatic driving or not selected by default, the automatic driving judgment module determines the switching timing, and the switching timing includes: the first switching timing: after the vehicle is started, the remote driving and automatic driving modes are preliminarily selected, and the preliminarily selected mode is the automatic driving mode, but the automatic driving function self-test fails; the second switching timing: during the automatic driving process, the automatic driving function is abnormal, and the system automatically switches the driving mode; the third switching timing: during the automatic driving process, the vehicle-mounted occupant manually switches the driving mode; the fourth switching timing: during the automatic driving process, the remote operator manually switches the driving mode; when the driving mode is remote driving, the remote driving judgment module determines the switching timing, and the switching timing includes: the fifth switching timing: during the remote driving process, the vehicle-mounted occupant manually switches the driving mode; the sixth switching timing: during the remote driving process, the remote operator manually switches the driving mode; starting from the first switching timing, in sequence to the sixth switching timing, the priority level of the switching timing gradually decreases. However, the different safety ranges of manual supervision and remote supervision are not considered, which leads to vehicle operation restrictions. For example, under manual supervision, the automatic driving is fault-free, but the remote driving fails, and the vehicle cannot be started. Summary of the invention

[0003] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method, system and medium for switching driving modes under different supervision modes, which can not only accurately switch the driving modes under different supervision modes, but also provide a safety redundancy mechanism between systems according to the two different supervision modes of in-vehicle supervision and remote supervision, thereby ensuring operational safety while ensuring operational efficiency.

[0004] In order to achieve the above-mentioned purpose and other related purposes, the technical solution provided by the present invention is as follows:

[0005] A driving mode switching method based on different supervision modes, the method comprising:

[0006] U1. The vehicle is driving on the road, collecting the vehicle's motion status data information in the in-vehicle supervision mode, and collecting the vehicle's motion status data information in the remote supervision mode;

[0007] U2. Based on the motion state data information of the vehicle in the in-vehicle supervision mode, the motion state of the vehicle is predicted by using a radial basis neural network algorithm based on time series analysis to obtain the predicted motion state data information of the vehicle in the in-vehicle supervision mode;

[0008] U3. Based on the motion state data information of the vehicle in the remote supervision mode, the motion state of the vehicle is predicted by using a deep separable neural network algorithm fused with the Spearman correlation coefficient to obtain the predicted motion state data information of the vehicle in the remote supervision mode;

[0009] U4. Based on the predicted vehicle motion state data information under the remote supervision mode and the predicted vehicle motion state data information under the in-vehicle supervision mode, a vehicle driving mode switching function Q under different modes is constructed, and the vehicle driving mode switching critical value is calculated to obtain the vehicle driving mode switching critical value data information.

[0010] Furthermore, the method further comprises:

[0011] U5. Based on the vehicle driving mode switching critical value data information, establish a vehicle driving mode switching function W,

[0012]

[0013] Among them, x is the vehicle driving mode switching critical value data information, α1 and α2 are the vehicle driving mode switching thresholds, the vehicle's driving mode is switched, and the vehicle's driving mode data information is output.

[0014] Furthermore, the constraint function f of the vehicle driving mode switching thresholds α1 and α2 is,

[0015] Among them, the value range of the constraint function f is (3,5).

[0016] Furthermore, in step U2, the use of a radial basis function neural network algorithm based on time series analysis to predict the motion state of the vehicle includes:

[0017] U21. Based on the vehicle's motion state data information in the in-vehicle supervision mode, establish a vehicle time series analysis function R,

[0018]

[0019] Among them, y is the motion state data information of the vehicle in the in-vehicle supervision mode, β1, β2 and β3 are the time series analysis factors of the vehicle motion state, which characterize the time series of the vehicle motion state and obtain the analyzed time series data information of the vehicle motion state;

[0020] U22. Input the analyzed time series data information of the vehicle motion state into the radial basis neural network for training and learning, and determine the radial basis function P of the network.

[0021]

[0022] Among them, z is the time series data information of the vehicle motion state after analysis, δ1, δ2 and δ3 are the extended constant parameters of the radial basis neural network, and the trained radial basis neural network is obtained;

[0023] U23. Based on the trained radial basis function neural network, the analyzed time series data information of the vehicle motion state is input, the vehicle motion state is predicted, and the predicted vehicle motion state data information under the in-vehicle supervision mode is obtained.

[0024] Furthermore, the extended constant parameters δ1, δ2 and δ3 of the radial basis neural network are,

[0025]

[0026] Among them, z is the time series data information of the vehicle motion state after analysis.

[0027] Furthermore, in step U3, the use of a deep separable neural network algorithm fused with the Spearman correlation coefficient to predict the motion state of the vehicle includes:

[0028] U31. Based on the vehicle motion state data information in the remote supervision mode, establish the Spearman correlation function S of the vehicle motion state,

[0029]

[0030] Among them, a is the motion state data information of the vehicle under remote supervision mode, γ1, γ2 and γ3 are the differential factors of the motion state of the vehicle, which characterize the correlation of the motion state of the vehicle and obtain the correlation data information of the motion state of the vehicle under remote supervision;

[0031] U32. Input the correlation data information of the motion state of the vehicle under remote supervision into the deep separable neural network for training and learning, and determine the neuron function G of the network,

[0032] Among them, b is the correlation data information of the motion state of the vehicle under remote supervision, η1, η2 and η3 are the weight constant parameters of the deep separable neural network, and the trained deep separable neural network is obtained;

[0033] U33. Based on the trained deep separable neural network, the motion state data information of the vehicle under remote supervision is input, the motion state of the vehicle is predicted, and the predicted motion state data information of the vehicle under remote supervision mode is obtained.

[0034] Furthermore, the constraints of the weight constant parameters η1, η2 and η3 of the deep separable neural network are:

[0035]

[0036] Furthermore, the vehicle driving mode switching function Q under different modes is:

[0037]

[0038]

[0039] Among them, c1 is the predicted motion state data information of the vehicle in the remote supervision mode, c2 is the predicted motion state data information of the vehicle in the in-vehicle supervision mode, and λ1, λ2 and λ3 are the switching factors of the vehicle driving mode.

[0040] In order to achieve the above-mentioned purpose and other related purposes, the present invention also provides a driving mode switching system based on different regulatory modes, including a computer device, which is programmed or configured to execute any one of the steps of the driving mode switching method based on different regulatory modes.

[0041] In order to achieve the above-mentioned purpose and other related purposes, the present invention also provides a computer-readable storage medium, which stores a computer program programmed or configured to execute any one of the driving mode switching methods based on different regulatory modes.

[0042] The present invention has the following positive effects:

[0043] 1. The present invention predicts the motion state of the vehicle by adopting a radial basis neural network algorithm based on time series analysis, and combines it with a deep separable neural network algorithm fused with the Spearman correlation coefficient to predict the motion state of the vehicle. It can not only accurately predict the motion state of the vehicle under different supervision modes and ensure the correctness of the vehicle driving mode switching, but also the whole process is highly robust and can adapt to different vehicle conditions, thereby improving the safety of vehicle driving.

[0044] 2. The present invention constructs a vehicle driving mode switching function Q under different modes and calculates the vehicle driving mode switching critical value. It can not only accurately switch the driving mode under different supervision modes, but also provide a safety redundancy mechanism between systems according to the two different supervision modes of in-vehicle supervision and remote supervision, thereby ensuring operational safety while ensuring operational efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0046] Figure 2 It is a schematic diagram of the flow of the radial basis neural network algorithm based on time series analysis of the present invention;

[0047] Figure 3 It is a schematic diagram of the flow of the deep separable neural network algorithm integrating the Spearman correlation coefficient of the present invention;

[0048] Figure 4 Schematic diagram of the system framework of the present invention. DETAILED DESCRIPTION

[0049] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0050] Example 1: Figure 1 As shown, a driving mode switching method based on different supervision modes includes:

[0051] U1. The vehicle is driving on the road, collecting the vehicle's motion status data information in the in-vehicle supervision mode, and collecting the vehicle's motion status data information in the remote supervision mode;

[0052] U2. Based on the motion state data information of the vehicle in the in-vehicle supervision mode, the motion state of the vehicle is predicted by using a radial basis neural network algorithm based on time series analysis to obtain the predicted motion state data information of the vehicle in the in-vehicle supervision mode;

[0053] U3. Based on the motion state data information of the vehicle in the remote supervision mode, the motion state of the vehicle is predicted by using a deep separable neural network algorithm fused with the Spearman correlation coefficient to obtain the predicted motion state data information of the vehicle in the remote supervision mode;

[0054] U4. Based on the predicted vehicle motion state data information under the remote supervision mode and the predicted vehicle motion state data information under the in-vehicle supervision mode, a vehicle driving mode switching function Q under different modes is constructed, and the vehicle driving mode switching critical value is calculated to obtain the vehicle driving mode switching critical value data information.

[0055] In this embodiment, the method further includes:

[0056] U5. Based on the vehicle driving mode switching critical value data information, establish a vehicle driving mode switching function W,

[0057]

[0058] Among them, x is the vehicle driving mode switching critical value data information, α1 and α2 are the vehicle driving mode switching thresholds, the vehicle's driving mode is switched, and the vehicle's driving mode data information is output.

[0059] In this embodiment, the constraint function f of the vehicle driving mode switching thresholds α1 and α2 is,

[0060]

[0061] Among them, the value range of the constraint function f is (3,5).

[0062] In this embodiment, if Figure 2 As shown, in step U2, the use of a radial basis function neural network algorithm based on time series analysis to predict the motion state of the vehicle includes:

[0063] U21. Based on the vehicle's motion state data information in the in-vehicle supervision mode, establish a vehicle time series analysis function R,

[0064]

[0065] Among them, y is the motion state data information of the vehicle in the in-vehicle supervision mode, β1, β2 and β3 are the time series analysis factors of the vehicle motion state, which characterize the time series of the vehicle motion state and obtain the analyzed time series data information of the vehicle motion state;

[0066] U22. Input the analyzed time series data information of the vehicle motion state into the radial basis neural network for training and learning, and determine the radial basis function P of the network.

[0067]

[0068] Among them, z is the time series data information of the vehicle motion state after analysis, δ1, δ2 and δ3 are the extended constant parameters of the radial basis neural network, and the trained radial basis neural network is obtained;

[0069] U23. Based on the trained radial basis function neural network, the analyzed time series data information of the vehicle motion state is input, the vehicle motion state is predicted, and the predicted vehicle motion state data information under the in-vehicle supervision mode is obtained.

[0070] In this embodiment, the extended constant parameters δ1, δ2 and δ3 of the radial basis neural network are,

[0071]

[0072] Among them, z is the time series data information of the vehicle motion state after analysis.

[0073] Embodiment 2: Based on the driving mode switching method under different supervision modes in Embodiment 1, the present invention is further illustrated and described below.

[0074] like Figure 1 As shown, a driving mode switching method based on different supervision modes includes:

[0075] U1. The vehicle is driving on the road, collecting the vehicle's motion status data information in the in-vehicle supervision mode, and collecting the vehicle's motion status data information in the remote supervision mode;

[0076] U2. Based on the motion state data information of the vehicle in the in-vehicle supervision mode, the motion state of the vehicle is predicted by using a radial basis neural network algorithm based on time series analysis to obtain the predicted motion state data information of the vehicle in the in-vehicle supervision mode;

[0077] U3. Based on the motion state data information of the vehicle in the remote supervision mode, the motion state of the vehicle is predicted by using a deep separable neural network algorithm fused with the Spearman correlation coefficient to obtain the predicted motion state data information of the vehicle in the remote supervision mode;

[0078] U4. Based on the predicted vehicle motion state data information under the remote supervision mode and the predicted vehicle motion state data information under the in-vehicle supervision mode, a vehicle driving mode switching function Q under different modes is constructed, and the vehicle driving mode switching critical value is calculated to obtain the vehicle driving mode switching critical value data information.

[0079] In this embodiment, if Figure 3As shown, in step U3, the use of the deep separable neural network algorithm fused with the Spearman correlation coefficient to predict the motion state of the vehicle includes:

[0080] U31. Based on the vehicle motion state data information in the remote supervision mode, establish the Spearman correlation function S of the vehicle motion state,

[0081]

[0082] Among them, a is the motion state data information of the vehicle under remote supervision mode, γ1, γ2 and γ3 are the differential factors of the motion state of the vehicle, which characterize the correlation of the motion state of the vehicle and obtain the correlation data information of the motion state of the vehicle under remote supervision;

[0083] U32. Input the correlation data information of the motion state of the vehicle under remote supervision into the deep separable neural network for training and learning, and determine the neuron function G of the network,

[0084] Among them, b is the correlation data information of the motion state of the vehicle under remote supervision, η1, η2 and η3 are the weight constant parameters of the deep separable neural network, and the trained deep separable neural network is obtained;

[0085] U33. Based on the trained deep separable neural network, the motion state data information of the vehicle under remote supervision is input, the motion state of the vehicle is predicted, and the predicted motion state data information of the vehicle under remote supervision mode is obtained.

[0086] In this embodiment, the constraints of the weight constant parameters η1, η2 and η3 of the deep separable neural network are:

[0087]

[0088] In this embodiment, the vehicle driving mode switching function Q under different modes is:

[0089] Among them, c1 is the predicted motion state data information of the vehicle in the remote supervision mode, c2 is the predicted motion state data information of the vehicle in the in-vehicle supervision mode, and λ1, λ2 and λ3 are the switching factors of the vehicle driving mode.

[0090] In this embodiment, the present invention provides a driving mode switching system based on different regulatory modes, including a computer device, which is programmed or configured to execute any one of the steps of the driving mode switching method based on different regulatory modes.

[0091] In this embodiment, if Figure 4 As shown, the remote human-computer interaction system functions as follows: manually configure the supervision mode, determine the supervision scope, in-vehicle supervision or remote supervision, which can be set on the remote test bench, remote platform, mobile terminal and other media;

[0092] Function of the remote platform main system: remotely manage the test bench, manage the vehicle, take over the vehicle, etc. through the platform;

[0093] The function of the remote test bench main system is to remotely take over the vehicle through the test bench;

[0094] Function of the remote platform safety redundancy system: to judge the fault condition of the overall platform system, including the communication condition inside and outside the system, and the fault condition of server software and hardware;

[0095] Function of the remote rack safety redundancy system: to determine the fault status of the overall rack system, including the communication status inside and outside the system, rack, server software and hardware fault status;

[0096] Function of the remote end communication system: It is installed on the remote device and is responsible for the internal communication of the remote end and the communication between the remote end and the vehicle end;

[0097] Function of the vehicle-mounted communication system: It is installed on the vehicle-mounted device and is responsible for the internal communication of the vehicle-mounted terminal and the communication between the vehicle-mounted terminal and the remote terminal;

[0098] Function of the vehicle-mounted remote driving safety redundancy system: to judge the fault conditions of the vehicle-mounted remote driving system, including the communication conditions inside and outside the system, the remote driving controller, and the software and hardware fault conditions of the remote driving sensor;

[0099] Function of the vehicle-mounted autonomous driving safety redundancy system: to determine the fault conditions of the vehicle-mounted autonomous driving system, including the communication conditions inside and outside the system, the software and hardware fault conditions of the autonomous driving controller and the autonomous driving sensor;

[0100] Function of the vehicle-mounted chassis safety redundancy system: to judge the fault conditions of the vehicle-mounted chassis domain, including the communication conditions inside and outside the system, the chassis domain controller, various chassis actuators, and the software and hardware fault conditions of the controller;

[0101] Function of AEB system: As a redundant system of the whole vehicle, it is responsible for providing the minimum emergency unit in the vehicle standby state, manual driving, remote driving and automatic driving;

[0102] The function of the vehicle-mounted remote driving system is to receive the sensing signals of the remote driving sensor and the control signals of the remote test bench main system to control the operation of the vehicle.

[0103] The function of the vehicle-mounted autonomous driving system is to receive the perception signals from the autonomous driving sensors, output control signals, and control the operation of the vehicle;

[0104] Chassis domain function: monitor the chassis condition, driving condition, and CAN communication of chassis-related parts.

[0105] In this embodiment, the present invention provides a computer-readable storage medium, on which is stored a computer program programmed or configured to execute any one of the driving mode switching methods based on different regulatory modes.

[0106] Any reference to memory, storage, database or other media used in the embodiments provided in the present application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM) and memory bus dynamic RAM (RDRAM).

[0107] In summary, the present invention can not only accurately switch driving modes under different supervision modes, but also provide a safety redundancy mechanism between systems according to the two different supervision modes of in-vehicle supervision and remote supervision, thereby ensuring operational safety while ensuring operational efficiency.

[0108] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A driving mode switching method based on different supervision modes, characterized in that: The method comprises: U1. The vehicle is driving on the road, collecting the vehicle's motion status data information in the in-vehicle supervision mode, and collecting the vehicle's motion status data information in the remote supervision mode; U2. Based on the motion state data information of the vehicle in the in-vehicle supervision mode, the motion state of the vehicle is predicted by using a radial basis neural network algorithm based on time series analysis to obtain the predicted motion state data information of the vehicle in the in-vehicle supervision mode; U3. Based on the motion state data information of the vehicle in the remote supervision mode, the motion state of the vehicle is predicted by using a deep separable neural network algorithm fused with the Spearman correlation coefficient to obtain the predicted motion state data information of the vehicle in the remote supervision mode; U4. Based on the predicted vehicle motion state data information under the remote supervision mode and the predicted vehicle motion state data information under the in-vehicle supervision mode, a vehicle driving mode switching function Q under different modes is constructed, and the vehicle driving mode switching critical value is calculated to obtain the vehicle driving mode switching critical value data information.

2. The driving mode switching method based on different supervision modes according to claim 1 is characterized in that: The method further comprises: U5. Based on the vehicle driving mode switching critical value data information, establish a vehicle driving mode switching function W, Among them, x is the vehicle driving mode switching critical value data information, α1 and α2 are the vehicle driving mode switching thresholds, the vehicle's driving mode is switched, and the vehicle's driving mode data information is output.

3. The driving mode switching method based on different supervision modes according to claim 2 is characterized in that: The constraint function f of the vehicle driving mode switching thresholds α1 and α2 is, Among them, the value range of the constraint function f is (3,5).

4. The driving mode switching method based on different supervision modes according to claim 1 is characterized in that: In step U2, the use of a radial basis function neural network algorithm based on time series analysis to predict the motion state of the vehicle includes: U21. Based on the vehicle's motion state data information in the in-vehicle supervision mode, establish a vehicle time series analysis function R, Among them, y is the motion state data information of the vehicle in the in-vehicle supervision mode, β1, β2 and β3 are the time series analysis factors of the vehicle motion state, which characterize the time series of the vehicle motion state and obtain the analyzed time series data information of the vehicle motion state; U22. Input the analyzed time series data information of the vehicle motion state into the radial basis neural network for training and learning, and determine the radial basis function P of the network. Among them, z is the time series data information of the vehicle motion state after analysis, δ1, δ2 and δ3 are the extended constant parameters of the radial basis neural network, and the trained radial basis neural network is obtained; U23. Based on the trained radial basis function neural network, the analyzed time series data information of the vehicle motion state is input, the vehicle motion state is predicted, and the predicted vehicle motion state data information under the in-vehicle supervision mode is obtained.

5. The driving mode switching method based on different supervision modes according to claim 4 is characterized in that: The extended constant parameters δ1, δ2 and δ3 of the radial basis neural network are, Among them, z is the time series data information of the vehicle motion state after analysis.

6. The driving mode switching method based on different supervision modes according to claim 1 is characterized in that: In step U3, the use of a deep separable neural network algorithm fused with the Spearman correlation coefficient to predict the motion state of the vehicle includes: U31. Based on the vehicle motion state data information in the remote supervision mode, establish the Spearman correlation function S of the vehicle motion state, Among them, a is the motion state data information of the vehicle under remote supervision mode, γ1, γ2 and γ3 are the differential factors of the motion state of the vehicle, which characterize the correlation of the motion state of the vehicle and obtain the correlation data information of the motion state of the vehicle under remote supervision; U32. Input the correlation data information of the motion state of the vehicle under remote supervision into the deep separable neural network for training and learning, and determine the neuron function G of the network, Among them, b is the correlation data information of the motion state of the vehicle under remote supervision, η1, η2 and η3 are the weight constant parameters of the deep separable neural network, and the trained deep separable neural network is obtained; U33. Based on the trained deep separable neural network, the motion state data information of the vehicle under remote supervision is input, the motion state of the vehicle is predicted, and the predicted motion state data information of the vehicle under remote supervision mode is obtained.

7. The driving mode switching method based on different supervision modes according to claim 6 is characterized in that: The constraints of the weight constant parameters η1, η2 and η3 of the deep separable neural network are:

8. The driving mode switching method based on different supervision modes according to claim 1 is characterized in that: The vehicle driving mode switching function Q under different modes is: Among them, c1 is the predicted motion state data information of the vehicle in the remote supervision mode, c2 is the predicted motion state data information of the vehicle in the in-vehicle supervision mode, and λ1, λ2 and λ3 are the switching factors of the vehicle driving mode.

9. A driving mode switching system based on different supervision modes, comprising a computer device, characterized in that: The computer device is programmed or configured to execute the steps of the driving mode switching method based on different supervision modes as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program that is programmed or configured to execute the driving mode switching method based on different supervision modes as described in any one of claims 1 to 8.

Images