Full-active intelligent chassis control system and method based on voice programming and AI decision

Abstract

The invention provides a full-active intelligent chassis control system and method based on voice programming and AI decision, and relates to the technical field of intelligent vehicle chassis control, and the system comprises the following modules: a voice collection and recognition module which carries out the semantic analysis of collected user voice; the demand programming module is used for setting control logic and parameters of each module of the chassis; the entertainment function module comprises a music rhythm extraction unit and a video motion matching unit; the mode selection module comprises a function specification mode and an entertainment mode and is used for selecting between the function specification mode and the entertainment mode; a CAN communication module; the AI decision controller is used for reasoning multi-source input information by using a deep learning model; an execution control module; the version management module is used for performing version management on the modification of the function specification file; according to the invention, rapid switching between a function specification mode and an entertainment mode can be realized through voice instructions, and brand new experience is provided for music rhythm and a 5D cinema function.

Description

Technical Field

[0001] This invention relates to the field of intelligent vehicle chassis control technology, and in particular to a fully active intelligent chassis control system and method based on voice programming and AI decision-making. Background Technology

[0002] In today's booming automotive industry, consumers' demands for cars have gone beyond basic transportation functions, placing higher demands on comfort, personalization, and safety in the driving experience. Traditional vehicle chassis control systems suffer from problems such as complex configurations, low personalization, slow response speeds, and a lack of deep integration of entertainment functions.

[0003] In existing technologies, firstly, chassis control and in-vehicle entertainment systems are independent of each other, making it impossible to achieve immersive experiences such as music rhythm and 5D cinema; secondly, the application of voice control technology in the chassis field is still immature, lacking a complete voice programming, version management, and entertainment function integration mechanism; at the same time, existing systems lack a safe switching mechanism between driving mode and entertainment mode; in addition, chassis control strategies are often fixed and complex, making them difficult for users to understand and modify, requiring professional engineers to program and calibrate, which limits the personalized customization and rapid iteration of chassis systems.

[0004] In addition, traditional chassis control systems generally use fixed-cycle algorithm control, which has a slow response speed and is difficult to meet the entertainment function requirements with high real-time requirements. Summary of the Invention

[0005] The purpose of this invention is to provide a fully active intelligent chassis control system and method based on voice programming and AI decision-making, thereby solving the technical problems existing in the prior art.

[0006] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows:

[0007] A fully active intelligent chassis control system and method based on voice programming and AI decision-making includes the following modules: a voice acquisition and recognition module, which converts acquired user voice into text using a voice recognition engine after filtering, noise reduction, and echo cancellation, and performs semantic parsing; a requirements programming module, which has a voice programming unit and manages multiple specification files, including JSON files defining input and output signals and boundary values, functional specification files, and standard format output JSON files; it is used to set the control logic and parameters of each chassis module; an entertainment function module, which includes a music rhythm extraction unit and a video motion matching unit; and a mode selection module, which includes functional specification modes and... The system includes an entertainment mode that receives voice commands from the voice acquisition and recognition module and selects between the functional specification mode and the entertainment mode; a CAN communication module for signal parsing and transmission with a 20ms cycle; an AI decision controller that uses a deep learning model to infer from multi-source input information and generates control strategies based on training data and rules; an execution control module, including an execution controller and actuators, that receives control commands from the AI ​​decision controller and controls corresponding components to perform corresponding actions; and a version management module that manages version modifications to the functional specification document, records the content, time, and modifier information of each modification, and supports version rollback operations.

[0008] Furthermore, it also includes a status feedback module, which collects various data of the vehicle in real time through the vehicle body status sensor and inputs the above feedback information into the AI ​​decision controller.

[0009] Furthermore, the entertainment function module has a music rhythm mode and a 5D cinema mode; the music rhythm mode processes audio signals through the music rhythm extraction unit; the 5D cinema mode processes video motion signals through the video motion matching unit.

[0010] Furthermore, the mode selection module establishes a mutually exclusive selection mechanism between the functional specification mode and the entertainment mode.

[0011] Furthermore, the functional specification file includes JSON files for the ECAS module, CDC module, fully active module, diagnostic module, arbitration module, and calibration variables.

[0012] Furthermore, the voice commands in the mode selection module include entering music mode, starting 5D cinema, and returning to driving mode.

[0013] Further, the process includes the following steps: S1: Voice command acquisition and parsing, acquiring user voice commands and performing semantic parsing; S2: Requirement conversion and parameter management, dynamically editing functional specification files and calibration variables based on the parsed requirements, or selecting entertainment function modes; simultaneously, the version management module manages modifications to the functional specification files; S3: Signal acquisition and mode integration, real-time cyclic parsing of sensor signals transmitted via the CAN bus, while simultaneously completing the mutual exclusion selection of functional specification files and entertainment functions; S4: Intelligent decision-making, inputting the acquired CAN sensor signals and the selected mode in step S3 into a deep learning model; based on the selected mode, the deep learning model fuses multi-source information to generate control commands; S5: Execution and closed-loop control, sending the AI ​​decision results to the execution controller via the CAN bus to drive the actuator to complete control actions; simultaneously acquiring real-time vehicle body status data and feeding it back to the AI ​​decision controller to dynamically adjust the control strategy.

[0014] Furthermore, when the selected mode is entertainment mode in step S4, the BPM rhythm of the audio signal is extracted first, motion template matching is performed on the video signal, and then the deep learning model fuses multi-source information to generate control commands.

[0015] Compared with the prior art, the present invention has the following beneficial effects:

[0016] (i) This invention allows users to quickly switch between the functional standard mode and the entertainment mode via voice commands without having to manually operate a complex interface or buttons. Furthermore, the music rhythm and 5D cinema functions provide a brand-new driving experience, deeply integrating chassis control with the aforementioned entertainment functions, bringing users an immersive driving experience and enhancing the convenience and interactivity of the user experience.

[0017] (ii) This invention uses voice programming to convert user adjustment commands into modifications to functional specification files, thereby changing chassis control parameters and enabling each user to create their own driving experience through personalized customization.

[0018] (iii) The functional specification mode and entertainment mode of this invention are strictly mutually exclusive, and the actuator movement range is limited in entertainment mode. Combined with vehicle speed detection and emergency exit mechanism, safety risks are avoided from the root and measures can be taken quickly to protect the vehicle and passengers even in extreme situations.

[0019] (iv) The present invention ensures the smoothness of entertainment functions through a 20ms control cycle. The millisecond-level control cycle enables the system to respond promptly to music rhythm and video action, ensuring the smooth operation of entertainment functions. Attached Figure Description

[0020] Figure 1 This is a diagram illustrating the overall architecture of the present invention;

[0021] Figure 2 This is a flowchart of the mode selection module of the present invention;

[0022] Figure 3 This is a flowchart of the entertainment function processing of the present invention;

[0023] Figure 4 This is the 20ms real-time control loop flowchart of the present invention; Detailed Implementation

[0024] To make the content of this invention easier to understand, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Identical components are represented by the same reference numerals. It should be noted that the terms "front," "rear," "left," "right," "up," and "down" used in the following description refer to directions in the accompanying drawings, while the terms "inner" and "outer" refer to directions toward or away from the geometric center of a specific component, respectively.

[0025] Example 1:

[0026] like Figure 1-4 As shown, this embodiment provides a fully active intelligent chassis control system and method based on voice programming and AI decision-making, including the following modules:

[0027] The voice acquisition and recognition module processes the acquired user voice through filtering, noise reduction, and echo cancellation. Then, it uses a voice recognition engine to convert the voice into text and performs semantic analysis to understand the user's intent. The voice recognition engine can use commercial models such as iFlytek. This module uses a 6-microphone array and a dedicated DSP chip to achieve noise reduction, echo cancellation, and voice wake-up, with a wake-up word recognition rate of ≥90%.

[0028] The requirements programming module, featuring a voice programming unit, manages multiple specification files, including functional specification files and calibration variables. Specifically, it includes JSON files defining input / output signals and boundary values, functional specification files, and standard format output JSON files. It is used to set the control logic and parameters for each chassis module. The functional specification files include JSON files for the ECAS module, CDC module, fully active module, diagnostic module, arbitration module, and calibration variables. ECAS refers to electronically controlled air suspension, and CDC refers to continuously variable damping. This ensures that the functional requirements of each system component are clearly defined and managed. During use, users can directly edit chassis control strategies and parameters via voice commands. When a user says, "I want the suspension to be stiffer," the system uses voice recognition and semantic analysis to translate the user's intent into modifications to the functional specification files and update the corresponding control parameters. The voice programming unit lowers the barrier to personalized chassis system customization, allowing users to adjust chassis performance in real time according to their preferences and driving needs, achieving a truly personalized driving experience.

[0029] The entertainment module includes a music rhythm extraction unit and a video motion matching unit, enabling music rhythm and 5D cinema functions. The entertainment module features a music rhythm mode and a 5D cinema mode, allowing users to select the appropriate mode to enhance their driving experience. The music rhythm mode processes audio signals through the music rhythm extraction unit. Specifically, in music rhythm mode, the system uses the music rhythm extraction unit to extract the rhythm of the in-vehicle audio signal and calculate rhythmic features such as BPM. The DSP used in this unit supports a 48kHz sampling rate, 24-bit precision audio processing, and real-time BPM detection accuracy ≤1BPM. The 5D cinema mode processes video motion signals through the video motion matching unit. Specifically, in 5D cinema mode, the system uses the video motion matching unit to match motion templates to the movie video played in the vehicle, identifying motion scenes and motion intensity in the video. This unit is equipped with an NVIDIA Jeston GPU module and supports 1080P video input. These two entertainment functions coexist with traditional chassis control functions, but are mutually exclusive through the mode selection module.

[0030] The mode selection module includes a functional standard mode and an entertainment mode. It receives voice commands from the voice acquisition and recognition module and selects between the two modes. The mode selection module establishes a mutual exclusion mechanism between the functional standard mode and the entertainment mode, ensuring that the system can only operate in one mode at any given time, thus preventing conflicts in control strategies between different modes. When the user selects to enter the entertainment mode, the system stops executing the regular chassis control strategy and instead executes the entertainment function control strategy. Conversely, when the user selects to return to the functional standard mode, the system resumes executing the regular chassis control strategy. The voice commands in the mode selection module include entering music mode, activating the 5D cinema, and returning to driving mode, etc., and users can speak the corresponding voice commands according to their needs.

[0031] The CAN communication module is used for signal parsing and transmission, with a working cycle of 20ms. This module is equipped with two high-speed CAN buses and supports CAN FD extended frames. It is used for data interaction between various actuators and sensors in the vehicle. It includes a CAN input signal parsing module and a CAN transmission signal loading module, which are responsible for parsing CAN signals from sensors in real time and loading decision results onto the CAN bus to send to the actuator controller, respectively. This ensures that key modules such as CAN input signal parsing, mode selection module, AI decision controller, and CAN transmission signal loading are executed in a fixed cycle of 20ms, thereby ensuring the real-time response performance of the control system.

[0032] The AI ​​decision controller utilizes a deep learning model to infer from multi-source input information and generates control strategies based on training data and rules. The deep learning model can be a DeepSeek model. The AI ​​decision controller receives real-time vehicle sensor data from the CAN input signal parsing module and current operating mode information from the mode selection module, and inputs this information into the DeepSeek model for inference. The DeepSeek model generates the optimal chassis control strategy or command based on the training data and rules. The sensors include four laser height sensors with an accuracy of ±1mm and a range of 0-300mm, three MEMS accelerometers with a range of ±2g and a bandwidth of 500Hz, and two wheel speed sensors with an accuracy of ±0.1km / h. This AI decision controller generates precise control commands for the vehicle's suspension, steering, braking, and other chassis systems through high-speed cyclic calculations.

[0033] The execution control module includes an execution controller and actuators. It receives control commands from the AI ​​decision controller and controls corresponding components to perform corresponding actions. The execution controller adopts a safety control strategy to limit and constrain the AI ​​decision results, ensuring that the output control commands are within a reasonable range and preventing vehicle loss of control due to AI decision errors or abnormal commands in entertainment mode. The actuators drive various components of the vehicle chassis, such as air springs, shock absorbers, and power steering motors, according to the instructions of the execution controller, to achieve active control of the vehicle's attitude and driving state. This module also has a functional safety mechanism that can automatically switch to a safety mode when the system malfunctions or is abnormal, ensuring the safety of the vehicle and passengers. Specifically, the actuators are configured with 9 air spring solenoid valves, 8 CDC solenoid valves, and 4 drive motors with a power of 50W and a speed range of 0-5000rpm.

[0034] The version management module manages modifications to functional specification documents, recording the content, time, and modifier information of each modification, and supports version rollback operations. When users modify chassis control strategies or parameters via voice commands, the system automatically generates a new version, saves the modification record, and prevents system instability caused by erroneous modifications. When needed, the version management module can be used to roll back to a previous stable version, ensuring the reliability and security of the system.

[0035] The status feedback module collects various vehicle data in real time through vehicle status sensors, including vehicle attitude, acceleration, and speed. This feedback information is then input into the AI ​​decision controller, which continuously adjusts its control strategy based on the feedback data to achieve the desired control objective. This module implements closed-loop control. When the user switches to entertainment mode, the status feedback module monitors the vehicle's vibration and attitude changes to ensure that music and 5D cinema actions remain within safe limits and do not affect vehicle stability. Through closed-loop control, the system can quickly respond to environmental changes and user commands, thereby achieving precise control and optimized adjustment of the vehicle chassis.

[0036] The method of using this invention includes the following steps:

[0037] S1: Voice command acquisition and parsing, acquiring user voice commands and performing semantic parsing to clarify the objectives for subsequent operations;

[0038] S2: Requirement transformation and parameter management. Based on the parsed requirements, the functional specification file and calibration variables are dynamically edited to achieve personalized chassis control; or entertainment function mode is selected; at the same time, the version management module manages the modification of the functional specification file to ensure that the modification is safe and traceable.

[0039] S3: Signal acquisition and mode integration, real-time cyclic analysis of sensor signals transmitted via CAN bus, and simultaneous selection of mutually exclusive functional specifications and entertainment functions to ensure single-mode operation of the system and avoid functional conflicts;

[0040] S4: Intelligent decision-making, inputting the collected CAN sensor signals and the selected mode in step S3 into the DeepSeek model; according to the selected mode, the DeepSeek model fuses multi-source information to generate control commands; if the entertainment mode is selected, first extract the BPM rhythm from the audio signal, perform motion template matching on the video signal, and then the DeepSeek model fuses multi-source information to generate control commands.

[0041] S5: Execution and closed-loop control. The AI ​​decision results are sent to the execution controller via the CAN bus to drive the actuators to complete the control actions. The drive actuators include air spring valves, CDC solenoid valves, etc. At the same time, real-time vehicle status data is collected and fed back to the AI ​​decision controller to dynamically adjust the control strategy to ensure control accuracy and system stability.

[0042] Example 2:

[0043] When a user needs to activate the music rhythm function, the user can say the voice command "Enter music mode and set the suspension rhythm matching intensity to 50%";

[0044] The voice acquisition and recognition module interprets the command as switching to music rhythm mode, setting the rhythm matching intensity to 0.5, and the mode selection module switches to entertainment mode. At this time, the system disables the loading of function specification files.

[0045] After the audio input system is activated, the music rhythm extraction unit analyzes the audio BPM at a period of 20ms and generates basic control parameters based on the rhythm matching intensity of 0.5.

[0046] The AI ​​decision controller combines BPM, rhythm intensity and vehicle status feedback to generate control commands for suspension lifting and damping adjustment.

[0047] After receiving CAN commands, the controller drives the air spring valve and CDC solenoid valve to make the suspension move rhythmically with the music; the vehicle body status feedback module collects sensor data in real time and adjusts the control accuracy in a closed loop to ensure that the movement range is within a safe range.

[0048] Example 3:

[0049] When a user needs to activate the 5D cinema function, the user can say the voice command "Open 5D cinema and load your favorite movie";

[0050] The voice acquisition and recognition module interprets the command as "Switch to 5D Cinema Mode and load the specified movie", and the mode selection module switches to entertainment mode, simultaneously calling the video processing unit;

[0051] The video processing unit performs frame-by-frame motion analysis on the specified video footage, extracts motion features such as vehicle acceleration, turning, and bumping, and generates motion templates.

[0052] The AI ​​decision controller combines the intensity of the motion template and the current state of the vehicle body to generate chassis attitude control commands. For example, in acceleration scenarios, the front suspension is compressed and the damping is stiffened; in cornering scenarios, the suspension on one side is compressed to simulate tilting.

[0053] When executing commands, the execution controller automatically triggers safety constraints; the vehicle body status feedback module monitors in real time to ensure that the attitude simulation is synchronized with the video and does not affect vehicle stability.

[0054] Example 4:

[0055] When the system is under security control mechanisms:

[0056] In functional specification mode, the entertainment function module is completely disabled; if an emergency braking signal is detected, the system immediately exits entertainment mode and switches to functional specification mode.

[0057] In entertainment mode, the suspension height adjustment range is limited, and the damping adjustment range is limited to 50% of the comfort and sport range to avoid extreme actions;

[0058] When the vehicle speed is greater than 40 km / h, the motion amplitude in the music rhythm mode is automatically halved, and the 5D cinema mode only allows slight posture simulation; when the vehicle speed is greater than 55 km / h, all entertainment modes are automatically exited and the system switches back to the standard function mode.

[0059] It supports voice command "emergency exit" or physical button triggering. The system exits entertainment mode and switches to functional specification mode within 20ms, while the actuator quickly restores to the default safe state.

[0060] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A fully active intelligent chassis control system and method based on voice programming and AI decision-making, characterized in that: The system comprises the following modules: A voice collection and recognition module, which converts the collected user voice into text through filtering, noise reduction, and echo cancellation, and performs semantic analysis; A requirement programming module, which has a voice programming unit, manages multiple Spec files, and contains input and output signal definitions and boundary value JSON files, functional specification files, and standard format output JSON files; It is used to set the control logic and parameters of each module of the chassis; An entertainment function module, which contains a music rhythm extraction unit and a video motion matching unit; A mode selection module, which includes a functional specification mode and an entertainment mode, receives voice instructions from the voice collection and recognition module, and selects between the functional specification mode and the entertainment mode; A CAN communication module, which is used for signal analysis and transmission, and has a working cycle of 20 ms; An AI decision controller, which uses a deep learning model to infer multiple source input information and generates a control strategy based on training data and rules; An execution control module, which includes an execution controller and an executor, receives control instructions from the AI decision controller, and controls corresponding components to perform corresponding actions; A version management module, which manages the version of the functional specification file, records the content, time, and modifier information of each modification, and supports version rollback operations.

2. The full active intelligent chassis control system based on voice programming and AI decision of claim 1, wherein: It also includes a state feedback module, which collects various data of the vehicle in real time through vehicle body state sensors, and inputs the above feedback information into the AI decision controller.

3. The full active intelligent chassis control system based on voice programming and AI decision of claim 1, wherein: The entertainment function module has a music rhythm mode and a 5D cinema mode; the music rhythm mode processes audio signals through the music rhythm extraction unit; the 5D cinema mode processes video action signals through the video motion matching unit.

4. The full active intelligent chassis control system based on voice programming and AI decision of claim 3, wherein: The mode selection module constructs a mutual exclusion selection mechanism between the functional specification mode and the entertainment mode.

5. The full active intelligent chassis control system based on voice programming and AI decision of claim 1, wherein: The functional specification file contains JSON files of ECAS modules, CDC modules, full-active modules, diagnosis modules, arbitration modules, and calibration variables.

6. The full active intelligent chassis control system based on voice programming and AI decision of claim 1, wherein: The voice instructions in the mode selection module include entering the music mode, starting the 5D cinema, and returning to the driving mode.

7. The full active intelligent chassis control system and method based on voice programming and AI decision according to any one of claims 1-6, characterized in that: The system comprises the following steps: S1: Voice instruction collection and analysis, collect user voice instructions and perform semantic analysis; S2: Requirement conversion and parameter management, dynamically edit the functional specification file and calibration variables according to the analyzed requirements, or select the entertainment function mode; at the same time, the version management module manages the modification of the functional specification file; S3: Signal collection and mode integration, real-time cyclically analyze the sensor signals transmitted by the CAN bus, and complete the mutual exclusion selection of the functional specification file and the entertainment function; S4: Intelligent decision-making, input the collected CAN sensor signals and the selected mode in step S3 into the deep learning model; generate control instructions based on the selected mode by fusing multiple source information from the deep learning model; S5: Perform closed-loop control, send the AI decision result to the execution controller through the CAN bus, drive the execution mechanism to complete the control action; at the same time, collect the real-time state data of the vehicle body and feed back to the AI decision controller to realize dynamic adjustment of the control strategy.

8. The full active intelligent chassis control system and method based on voice programming and AI decision according to claim 7, characterized in that: When the selected mode in the step S4 is the entertainment mode, BPM rhythm is extracted from the audio signal, motion template matching is performed on the video signal, and then a deep learning model is used to fuse multi-source information to generate a control instruction.

Images