Integrated control system for master-slave following self-balancing two-wheeled vehicle

Abstract

The application relates to the technical field of control, and discloses an integrated control system for master-slave following of a self-balancing two-wheeled vehicle, which comprises a master vehicle control end, a slave vehicle control end and a communication module. The master vehicle control end collects biomechanics and kinematics data by means of an intention analysis module to obtain driving intention features, and maps the driving intention features into a prediction control vector containing a future time window and sends the prediction control vector to the slave vehicle through a prediction instruction generation module; the slave vehicle control end estimates dynamics boundaries based on suspension travel and load by means of an envelope verification module and compares the vector, and a cooperative following execution module selects a strategy according to a verification result: when the vector is not out of limit, posture synchronization is executed; when the vector is out of limit, a trajectory compensation or speed-radius coupling adjustment strategy is executed. Through intention prediction and dynamics boundary verification, the integrated control of the master-slave following of the self-balancing two-wheeled vehicle realizes zero-lag response and active anti-rollover, and the safety and stability of formation driving are ensured.

Description

Technical Field

[0001] This invention relates to the field of control technology, specifically to an integrated control system for a self-balancing two-wheeled vehicle that uses a master-slave following mechanism. Background Technology

[0002] With the development of short-distance transportation and commercial logistics, self-balancing two-wheeled vehicles are increasingly being used in platooning applications in home and commercial settings. This application model typically involves a mobile, manned master vehicle guiding multiple slave vehicles in coordinated movement. This requires the system to not only maintain platooning formation in complex environments but also to ensure the load transport and driving safety of each vehicle.

[0003] Existing control technologies for master-slave following primarily rely on vision or ultra-wideband sensors to detect the relative position and distance of the master vehicle in real time, employing a reactive following strategy based on position error feedback. After detecting displacement or speed deviations in the master vehicle, the slave vehicle uses PID control algorithms to accelerate or steer its wheels to eliminate spatial errors with the master vehicle. This approach is commonly used in constant speed or low-speed steady-state driving.

[0004] However, due to the non-minimum phase characteristic of self-balancing two-wheeled vehicles, acceleration can only be generated by changing the vehicle's posture first. Traditional reactive control lacks the ability to perceive the rider's intentions and cannot predict changes in the master vehicle's acceleration in advance. This results in a persistent phase lag in the slave vehicle's posture adjustment, making it prone to losing target when the master vehicle accelerates rapidly or turns quickly. Furthermore, existing technologies typically do not consider the dynamic limits of the slave vehicle under different loads. Blindly replicating the master vehicle's actions can easily cause the slave vehicle to exceed its stability boundaries and tip over, or fall out of communication range due to an excessively large turning radius. This makes it difficult to meet the requirements of high dynamic coordination and active safety for integrated control of self-balancing two-wheeled vehicles that require master-slave following.

[0005] Therefore, this invention proposes an integrated control system for self-balancing two-wheeled vehicles that supports master-slave following, in order to address the shortcomings of existing technologies. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides an integrated control system for self-balancing two-wheeled vehicles that enables master-slave following. This system solves the problems of existing technologies, which lack perception and prediction of the rider's intentions and do not consider the dynamic limits of the slave vehicle, resulting in delayed posture response, easy tipping and instability, and loss of communication coverage during the following process.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an integrated control system for a self-balancing two-wheeled vehicle with master-slave following capability, comprising: a master vehicle control terminal, a slave vehicle control terminal, and a communication module; The main vehicle control terminal includes an intent parsing module and a prediction instruction generation module; The intent parsing module collects the biomechanical signals applied by the rider to the master vehicle and the kinematic data of the master vehicle, parses the driving intent features and uses the kinematic data to perform temporal consistency verification to obtain the verified and valid driving intent features; the prediction instruction generation module maps the verified and valid driving intent features into a predictive control vector containing future time window information, and sends it to the slave vehicle control terminal through the communication module. The slave vehicle control unit includes an envelope verification module and a cooperative following execution module. The envelope verification module receives the predicted control vector, estimates the physical load parameters of the slave vehicle based on the current suspension travel and load state, calculates the dynamic stability boundary, and compares the predicted control vector with the dynamic stability boundary to obtain the verification result. The cooperative following execution module selects a strategy based on the verification result: if the verification result does not exceed the limit, it executes the attitude synchronization strategy; if it exceeds the limit, it executes the trajectory compensation strategy or the speed radius coupling adjustment strategy.

[0008] Preferably, the step of parsing the driving intention features and performing temporal consistency verification on them using the kinematic data to obtain valid driving intention features includes: The intent parsing module performs spatial synthesis processing on the collected seat pressure distribution data and foot pedal pressure data to calculate the change in the position of the pressure center relative to the geometric center of the main vehicle seat. By combining the collected steering torque input, the pressure center position change and the steering torque input are transformed into a longitudinal intention component representing acceleration and deceleration demand and a lateral intention component representing steering demand using a multi-parameter fusion mapping model, which serve as the driving intention features. Set a verification time window, and calculate the correlation energy between the longitudinal intention component, the lateral intention component and the actual longitudinal acceleration and actual roll rate of the main vehicle within the verification time window; The correlation energy is compared with the confidence threshold. If the correlation energy is greater than the confidence threshold, the driving intention feature is determined to be valid, and the verified valid driving intention feature is obtained.

[0009] Preferably, the step of the prediction instruction generation module mapping the verified and valid driving intention features into a predictive control vector containing future time window information includes: The longitudinal and lateral intention components in the verified and valid driving intention features are mapped to expected acceleration and expected path curvature using the main vehicle dynamic response model. The target time for the action to take effect is calculated. The target time is obtained by superimposing the current system time, the transmission delay of the communication module, the mechanical response time of the actuator of the vehicle, and the forward prediction time associated with the speed. Determine the communication coverage radius constraint parameters based on the physical characteristics of the communication module; The expected acceleration, the expected path curvature, the target time, and the communication coverage radius constraint parameters are encapsulated into the predictive control vector.

[0010] Preferably, the step of the envelope verification module estimating the physical load parameters of the vehicle and calculating the dynamic stability boundary based on the current suspension travel and load state of the vehicle includes: Calculate the current total mass of the vehicle based on the current suspension travel and suspension spring stiffness characteristics of the vehicle. The current center of gravity height of the vehicle is calculated based on the vehicle's unloaded center of gravity height, the preset height of the cargo area, and the vehicle's current total mass. The anti-rollover lateral acceleration threshold is calculated based on the current center of gravity height of the vehicle and the wheel track width. An elliptical envelope region with longitudinal acceleration and lateral acceleration as variables is constructed as the dynamic stability boundary. The boundary of the elliptical envelope region is defined by the longitudinal acceleration limit corresponding to the road friction coefficient and the anti-rollover lateral acceleration threshold.

[0011] Preferably, the step of the envelope verification module comparing the predicted control vector with the dynamic stability boundary to obtain the verification result includes: The requested lateral acceleration is calculated based on the expected path curvature in the predicted control vector and the vehicle's current speed. Determine whether the state point formed by the expected acceleration in the predicted control vector and the requested lateral acceleration is located within the elliptical envelope region; If the state point is located within the elliptical envelope region and satisfies the preset safety margin coefficient, then the verification result is determined to be within the limit; Otherwise, the verification result is determined to be out of bounds.

[0012] Preferably, the step of the cooperative following execution module executing the attitude synchronization strategy includes: The preset pitch angle to counteract the longitudinal inertial force is calculated based on the expected acceleration in the predicted control vector, and the preset roll angle to counteract the centrifugal force is calculated based on the expected path curvature in the predicted control vector and the current vehicle speed. Generate an interpolated trajectory from the current attitude to the preset pitch angle and the preset roll angle; drive the self-balancing mechanism of the slave vehicle to output a feedforward torque so that the slave vehicle body completes attitude tilting before the target time specified by the predicted control vector; The driving mechanism of the vehicle outputs driving torque according to the expected acceleration.

[0013] Preferably, when the verification result is an over-limit, the cooperative follow-up execution module is specifically used for: The minimum safe turning radius of the vehicle at present is calculated using the anti-rollover lateral acceleration threshold and the safety margin coefficient; The curvature of the planned compensation path corresponds to the minimum safe turning radius and is tangent to the trajectory of the master vehicle; the master-slave distance is predicted after the slave vehicle travels along the compensation path; The master-slave vehicle distance is compared with the communication coverage radius constraint parameter in the predictive control vector. If the master-slave vehicle distance is less than or equal to the communication coverage radius constraint parameter, the trajectory compensation strategy is selected to be executed. If the distance between the master and slave vehicles is greater than the communication coverage radius constraint parameter, then the speed radius coupling adjustment strategy is selected to be executed.

[0014] Preferably, the step of the collaborative following execution module executing the trajectory compensation strategy includes: Drive the vehicle's driving mechanism to maintain the current speed; The self-balancing mechanism of the driven vehicle is coordinated with the steering to control the driven vehicle to travel along the compensation path, thereby creating a lateral offset relative to the master vehicle.

[0015] Preferably, the step of the cooperative following execution module executing the speed radius coupling adjustment strategy includes: Based on the communication coverage radius constraint parameters, the anti-rollover lateral acceleration threshold, and the safety margin coefficient, the maximum allowable coupling speed is calculated in reverse. Generate a braking command to drive the vehicle's driving mechanism to decelerate to the maximum permissible coupling speed; The generated steering command drives the self-balancing mechanism of the slave vehicle to adjust the roll angle, controlling the slave vehicle to travel with the turning radius corresponding to the communication coverage radius constraint parameter.

[0016] Preferably, the communication module is used to establish a bidirectional data link between the master vehicle control terminal and the slave vehicle control terminal using a point-to-point wireless technology with ultra-wideband wireless communication protocol or WIFI direct connection.

[0017] This invention provides an integrated control system for a self-balancing two-wheeled vehicle with master-slave following capability. It offers the following advantages: 1. This invention deeply integrates the rider's biomechanical signals and kinematic data through an intent parsing module to construct a predictive control vector containing future time window information. Human behavior prediction is completed on the master vehicle and transmitted via the communication module. This mechanism enables systems used for master-slave following to overcome the phase lag of traditional feedback control, allowing the self-balancing two-wheeled vehicle to anticipate changes in the master vehicle's acceleration and adjust its posture in advance on the slave vehicle. This achieves millisecond-level motion synchronization at the integrated control level, ensuring rapid response from the slave vehicle when the master vehicle changes its center of gravity.

[0018] 2. This invention uses an envelope verification module to read the suspension travel in real time to estimate physical load parameters and construct a dynamic dynamic stability boundary. Before executing a following action, it pre-determines whether the command exceeds the physical limits. This design ensures that in master-slave following mode, self-balancing two-wheeled vehicles with different load states can intelligently switch strategies according to their own capabilities. When facing the risk of rollover, they automatically perform trajectory compensation to adjust their relative position, thereby constructing an active risk avoidance mechanism in the integrated control system and greatly improving the driving safety of multi-vehicle platoons.

[0019] 3. This invention introduces dual constraints of communication coverage radius and dynamic boundary through a collaborative following execution module. When the predicted distance is about to exceed the limit, a speed radius coupling adjustment strategy is triggered to actively limit the speed of the follower vehicle in exchange for a smaller turning radius. This method solves the link interruption problem caused by differences in maneuverability in master-slave following scenarios, ensuring that the self-balancing two-wheeled vehicle can always remain within the effective communication range. Through precise integrated control logic, a stable following distance is maintained, meeting the continuous and reliable platooning requirements in home or commercial accompanying scenarios. Attached Figure Description

[0020] Figure 1 This is a diagram of the integrated control system architecture for a self-balancing two-wheeled vehicle using master-slave following for the present invention. Figure 2 This is a flowchart of the logic processing of the envelope verification module of the present invention; Figure 3 This is a flowchart of the integrated control method for a self-balancing two-wheeled vehicle using master-slave following according to the present invention.

[0021] Among them, 100 is the master vehicle control terminal; 110 is the intent parsing module; 120 is the prediction instruction generation module; 200 is the slave vehicle control terminal; 210 is the envelope verification module; 220 is the cooperative following execution module; and 300 is the communication module. Detailed Implementation

[0022] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] See attached document Figure 1This invention provides an integrated control system for self-balancing two-wheeled vehicles with master-slave following capability. This system is applied in platooning scenarios comprising at least one manned master vehicle and at least one unmanned follower vehicle. The integrated control system includes a master vehicle control terminal 100, a slave vehicle control terminal 200, and a communication module 300 connecting the master vehicle control terminal 100 and the slave vehicle control terminal 200. The master vehicle control terminal 100 is deployed on a master vehicle platform with high mobility and variable center of gravity capability, while the slave vehicle control terminal 200 is deployed on a slave vehicle platform with low mobility and a fixed center of gravity. The communication module 300 establishes a bidirectional data link between the master vehicle control terminal 100 and the slave vehicle control terminal 200 using a point-to-point wireless technology employing an ultra-wideband wireless communication protocol or a direct Wi-Fi connection.

[0024] The main vehicle control terminal 100 includes an intent parsing module 110 and a prediction instruction generation module 120.

[0025] The intent parsing module 110 is configured to collect biomechanical signals applied by the rider to the master vehicle and kinematic state data of the master vehicle itself in real time. Based on the biomechanical signals, it calculates the change in the position of the pressure center and the input of steering torque to parse the driving intent features. It also uses the kinematic state data of the master vehicle itself to perform temporal consistency verification on the driving intent features to obtain valid driving intent features.

[0026] The prediction instruction generation module 120 is configured to receive a verified and valid driving intention feature, map the driving intention feature into a prediction control vector containing future time window information, and send the prediction control vector to the slave control terminal 200 through the communication module 300.

[0027] The vehicle control terminal 200 includes an envelope verification module 210 and a cooperative follow-up execution module 220.

[0028] See attached document Figure 2 The envelope verification module 210 is configured to receive the predictive control vector, read the current suspension travel and load status of the slave vehicle to estimate the physical load parameters of the slave vehicle, calculate the dynamic stability boundary of the slave vehicle based on this, compare the predictive control vector with the dynamic stability boundary of the slave vehicle, determine whether the action command requested by the master vehicle exceeds the current physical limit of the slave vehicle, and output the verification result to the cooperative following execution module 220.

[0029] The collaborative follow-up execution module 220 is configured to select an execution strategy based on the verification result; when the verification result is within the limit, it controls the slave vehicle to execute the attitude synchronization strategy; when the verification result is beyond the limit, it controls the slave vehicle to execute the trajectory compensation strategy or the speed radius coupling adjustment strategy, driving the self-balancing mechanism and the driving mechanism of the slave vehicle to move.

[0030] See attached document Figure 3This invention provides an integrated control method for a self-balancing two-wheeled vehicle using a master-slave following mechanism. The method is based on an integrated control system for a self-balancing two-wheeled vehicle using a master-slave following mechanism and includes the following steps: S10, the intent parsing module 110 in the master vehicle control terminal 100 collects the biomechanical signals applied by the rider to the master vehicle and the kinematic state data of the master vehicle itself in real time. S20, the intent parsing module 110 calculates the change in the position of the pressure center and the input of the steering torque based on the biomechanical signal to parse the driving intent features, and uses the kinematic state data of the main vehicle itself to perform temporal consistency verification on the driving intent features to obtain the verified valid driving intent features. S30, the prediction instruction generation module 120 receives the verified and valid driving intention feature, maps the driving intention feature into a prediction control vector containing future time window information, and sends the prediction control vector to the slave control terminal 200 through the communication module 300. S40, the envelope verification module 210 in the vehicle control terminal 200 receives the predicted control vector, reads the current suspension travel and load status of the vehicle to estimate the physical load parameters of the vehicle, and calculates the dynamic stability boundary of the vehicle accordingly. S50, the envelope verification module 210 compares the predicted control vector with the dynamic stability boundary of the slave vehicle to determine whether the action command requested by the master vehicle exceeds the current physical limit of the slave vehicle, and outputs the verification result to the cooperative following execution module 220; S60, the collaborative following execution module 220 selects an execution strategy based on the verification result; when the verification result is within the limit, the collaborative following execution module 220 controls the slave vehicle to execute the attitude synchronization strategy; when the verification result is beyond the limit, the collaborative following execution module 220 controls the slave vehicle to execute the trajectory compensation strategy or the speed radius coupling adjustment strategy.

[0031] To further clarify the technical details of the system and method in this embodiment, the specific implementation methods, principles and technical contents of each module and method will be described in detail below.

[0032] See attached document Figure 1 and Figure 3 The intent parsing module 110 performs logical processing aimed at extracting real driving commands from complex human-machine interaction signals. The intent parsing module 110 collects biomechanical signals applied by the rider to the vehicle and kinematic state data of the vehicle itself in real time via a physical interface. Biomechanical signals include seat pressure distribution data from a thin-film pressure sensor, pedal pressure data from a resistance strain gauge, and steering torque input from a handlebar torque sensor. Kinematic state data includes the vehicle's pitch rate, roll rate, and triaxial acceleration collected by an inertial measurement unit.

[0033] The intent parsing module 110 performs spatial synthesis processing on the acquired biomechanical signals to calculate the change in the pressure center position. For signals containing... The sensor unit's seat pressure array and left and right foot pedals are used to establish a base coordinate system with the geometric center of the main vehicle seat as the origin. The weighted average algorithm is then used to calculate the coordinates of the entire vehicle's pressure center. The calculation process is based on the following formula: ; in, and These are the real-time position coordinates of the pressure center in the longitudinal and transverse directions, respectively. For the first in the seat cushion array The pressure value detected by each sensing unit These are the geometric coordinates of the unit. and These are the vertical pressure values ​​detected by the left and right pedals, respectively. and These are the coordinate constants of the centers of the left and right foot pedals, respectively; This is the sum of the total pressures detected.

[0034] After obtaining the change in the pressure center position, the intent parsing module 110, in conjunction with the steering torque input, parses the driving intent features. This parsing process employs a multi-parameter fusion mapping model, transforming the displacement of the body's center of gravity and the torque applied by the hands into a standardized control intent vector. The driving intent features include a longitudinal intent component representing acceleration and deceleration demands. and the lateral intention component representing the shift in demand The calculation logic is based on the following formula: ; in, and This serves as a reference value for the center of pressure of a cyclist in a static, balanced state. Input the value for the collected steering torque; This is the positional proportional gain coefficient, used to quantify the contribution of body tilt amplitude to the intensity of intent; This is the differential gain coefficient, used to capture the dynamic trend of rapid changes in the center of gravity; This is the steering torque weighting coefficient. Using this formula, the intent parsing module 110 quantifies the rider's forward acceleration, backward braking, side-leaning turns, and reverse steering actions into numerical driving intent features.

[0035] The intent parsing module 110 then uses the vehicle's own kinematic state data to perform a temporal consistency check on the aforementioned driving intent features, in order to eliminate false trigger signals caused by the rider adjusting their posture or unconscious swaying. The temporal consistency check is based on the causal delay principle of action and response, meaning that an effective human control action must inevitably trigger a vehicle dynamics response that conforms to physical laws within a specific time window. The intent parsing module 110 sets a verification time window. (e.g., 0.2 to 0.5 seconds), and within this window, calculate the correlation energy between driving intention features and actual kinematic state data. The calculation is based on the following formula: ; in, This represents the actual roll velocity of the vehicle body. This is the actual longitudinal acceleration of the vehicle body; For integration variables. The intent parsing module 110 will calculate the... Compared with the preset confidence threshold (For example, 0.8) is used for comparison. When When this occurs, it indicates that the current human action has been effectively converted into vehicle movement, and is determined to be a genuine driving command. The system then outputs a verified and valid driving intention feature. If the human body movement does not elicit the expected dynamic response or the response direction is contrary to the expected direction, it is judged as interference noise, and the system suppresses the driving intention characteristics at that moment. For filtering and preprocessing of the inertial measurement unit data, those skilled in the art can use conventional methods such as Kalman filtering, which are well-known technologies in the field and will not be elaborated upon here. Through the above processing, the intention parsing module 110 ensures that the output signal accurately reflects the rider's true control purpose.

[0036] See attached document Figure 1 and Figure 3 The prediction instruction generation module 120 is connected to the data output terminal of the intent parsing module 110 and is configured to perform the task of converting intent features into physical control instructions. The prediction instruction generation module 120 receives verified and valid driving intent features and maps them into predictive control vectors containing future time window information through dynamic mapping and temporal compensation processing.

[0037] The prediction instruction generation module 120 first reads the current vehicle speed signal of the master vehicle. and the longitudinal intent component in the verified valid driving intent features. and lateral intention components The prediction instruction generation module 120 uses a preset vehicle dynamics response model to transform dimensionless intentional features into expected motion parameters with clear physical meaning. This mapping process calculates the expected acceleration. and expected path curvature The calculation is based on the following formula: ; in, This is the longitudinal acceleration mapping gain coefficient, used to define the linear relationship between the intent intensity and the acceleration value; This is the steering response gain coefficient; A minimum speed threshold (e.g., 0.1 m / s) is set to prevent the denominator from being zero. Using this formula, the prediction command generation module 120 determines the expected motion state of the main vehicle at future moments.

[0038] After determining the expected motion parameters, the prediction command generation module 120 constructs a predictive control vector incorporating future time window information. This future time window information is used to compensate for communication transmission delays between the master and slave vehicles and the mechanical response lag of the slave vehicle's actuators, ensuring that the slave vehicle can maintain synchronization with the master vehicle's movements or make attitude adjustments in advance. The prediction command generation module 120 calculates the target time when the action takes effect. The calculation is based on the following formula: ; in, The current clock timestamp; The average transmission delay constant of the communication module 300 (e.g., 5ms); The mechanical time constant (e.g., 20 ms) required for the actuator to establish a steady-state response. The forward prediction coefficients related to speed are given. It clearly specifies the exact time at which the vehicle should complete the task. and The defined actions thus eliminate phase lag in traditional follower systems.

[0039] The prediction instruction generation module 120 further combines the physical characteristics of the communication module 300 to determine the communication coverage radius constraint parameters. This radius constraint parameter is set based on the effective ranging range of UWB (Ultra-Wideband) signals or WIFI Direct (point-to-point wireless technology) signals (e.g., 30 meters), and is used to limit the maximum following degree of freedom of the vehicle. The prediction command generation module 120 encapsulates the above parameters into a predictive control vector. Its mathematical expression is: ; After construction, the prediction command generation module 120 sends the predicted control vector to the slave vehicle control terminal 200 via the communication module 300. The communication module 300 is configured to establish a low-latency data link between the master vehicle and the slave vehicle, encoding the digitized predictive control vector into a radio frequency signal for broadcast. For the data link establishment protocol and underlying modulation and demodulation technology, those skilled in the art can use standard protocols such as IEEE 802.15.4z or Wi-Fi Direct, which are well-known technologies in the field and will not be elaborated further here. Through the above process, the slave vehicle control terminal 200 obtains a future action command containing clear three-dimensional information of time, space, and constraints, rather than just a snapshot of the current state.

[0040] See attached document Figures 1-3 The logic processing performed by the envelope verification module 210 is designed to ensure the physical safety of the follower vehicle during the following process, preventing rollover or loss of control caused by blindly replicating the high-maneuverability actions of the master vehicle. The envelope verification module 210 is configured to receive the predictive control vector and read the current suspension travel and load status of the follower vehicle to estimate the physical load parameters of the follower vehicle.

[0041] Envelope verification module 210 reads the values ​​of the linear displacement sensor installed on the slave vehicle suspension mechanism through the internal bus interface to obtain the current suspension travel of the slave vehicle. Load condition refers to the current load mass and distribution of the vehicle, which is characterized by the compression change in suspension travel. The envelope verification module 210 estimates the vehicle's physical load parameters based on the read suspension travel data and pre-calibrated suspension spring stiffness characteristics. The vehicle's physical load parameters include the vehicle's current total mass. and the current center of gravity height of the car The estimation process is based on the following formula: ; in, The basic mass of the vehicle when it is unloaded; This is the stiffness coefficient of the elastic element in the suspension mechanism; It is the constant of gravitational acceleration; This refers to the height of the vehicle's center of gravity when it is unloaded. This is a preset height value (e.g., 0.5 meters) of the geometric center of the cargo area relative to the ground. Through this calculation, the envelope verification module 210 obtains inertial parameters that reflect the current actual load conditions.

[0042] After obtaining the vehicle's physical load parameters, the envelope verification module 210 calculates the vehicle's dynamic stability boundary. The vehicle's dynamic stability boundary refers to the acceleration limit range under the current load and road surface conditions that allows the vehicle to maintain stable driving without skidding or overturning. The envelope verification module 210 constructs this boundary as an elliptical envelope region with longitudinal and lateral acceleration as variables. Its definition is based on the following formula: ; in, For longitudinal acceleration variables; It is the lateral acceleration variable; It is a two-dimensional real number space; The road surface friction coefficient (can be set to a conservative value such as 0.6). The lateral acceleration threshold for rollover resistance is calculated as follows: , Let be the wheelbase width of the vehicle. This elliptical envelope region visually describes the current physical limits of the vehicle.

[0043] The envelope verification module 210 then compares the received predicted control vector with the slave vehicle's dynamic stability boundary to determine whether the master vehicle's requested action command exceeds the slave vehicle's current physical limits. The envelope verification module 210 parses the expected acceleration in the predicted control vector. and expected path curvature Combined with the vehicle's current real-time driving speed Calculate the requested lateral acceleration required by the vehicle to execute the instruction. The calculation formula is The envelope verification module 210 verifies the data pairs. Does the formed state point lie within the elliptical envelope region? Within, and generate verification results. The judgment logic is based on the following formula: ; in, A preset safety margin coefficient (e.g., 0.85) is used to reserve a certain amount of control leeway. When the verification result is within the limit, it indicates that the master vehicle's action command is within the slave vehicle's capability range; when the verification result is beyond the limit, it indicates that the master vehicle's action command exceeds the slave vehicle's current physical limits. The envelope verification module 210 outputs the verification result to the cooperative following execution module 220 as the basis for subsequent decision-making on selecting an attitude synchronization strategy or a trajectory compensation strategy.

[0044] See attached document Figure 1 and Figure 3The logical processing of the cooperative following execution module 220 aims to transform the safety-verified decision into specific physical drive signals, driving the vehicle's self-balancing and driving mechanisms. The cooperative following execution module 220 is configured to select an execution strategy based on the verification results.

[0045] The collaborative follow-up execution module 220 receives the verification result from the envelope verification module 210. And the predictive control vector from the predictive command generation module 120. The cooperative follow execution module 220 judges the state of the verification result. When the verification result is within the limit, the cooperative follow execution module 220 controls the slave vehicle to execute the attitude synchronization strategy. This strategy uses the time window information in the predictive control vector to solve the dynamic lag problem of the self-balancing vehicle under the characteristics of a non-minimum phase system.

[0046] When executing the attitude synchronization strategy, the cooperative follow execution module 220 reads the expected acceleration from the predictive control vector. Expected path curvature and the target time when the action takes effect The cooperative follow-up execution module 220 calculates the preset attitude angles (e.g., 5 degrees) required for the vehicle to maintain dynamic balance, including a preset pitch angle to counteract longitudinal inertial forces. and the preset tilt angle to counteract centrifugal force (For example, 8 degrees). The calculation is based on the following formula: ; in, The current speed of the vehicle; The acceleration due to gravity. The cooperative follow execution module 220 generates a path that smoothly transitions from the current attitude to the target attitude. The interpolated trajectory is used to drive the vehicle's self-balancing mechanism (such as a reaction flywheel or torque gyroscope) to output feedforward torque, so that the vehicle body... The tilting motion is completed before the designated time. Subsequently, the coordinating follow-up execution module 220 drives the vehicle's driving mechanism (e.g., wheel hub motors) to accelerate as expected. Output drive torque. Through this timing control, the vehicle utilizes the gravitational component to counteract the inertial impact during acceleration or steering.

[0047] When the verification result exceeds the limit, the cooperative following execution module 220 controls the slave vehicle to execute a trajectory compensation strategy or a speed radius coupling adjustment strategy. At this time, the cooperative following execution module 220 first reads the anti-rollover lateral acceleration threshold calculated by the envelope verification module 210. And combined with the safety margin factor (e.g., 0.85), calculate the minimum safe turning radius from the current position of the vehicle. The calculation formula is The collaborative follow-up execution module 220 is based on... Plan a compensation path tangent to the main vehicle's trajectory, the curvature of which is... .

[0048] The collaborative following execution module 220 further predicts the master-slave distance after traveling along the compensation path. and will Communication coverage radius constraint parameters in predictive control vectors Perform a comparison. If... The collaborative following execution module 220 determines the execution trajectory compensation strategy, drives the driving mechanism to maintain the current speed, and drives the self-balancing mechanism to cooperate with the steering, so that the follower vehicle travels along the compensation path, thereby forming a lateral orientation offset relative to the master vehicle in space.

[0049] like This indicates that simply increasing the turning radius will lead to a communication link interruption. The cooperative following execution module 220 determines a speed-radius coupling adjustment strategy. This strategy achieves a smaller safe turning radius by actively reducing the speed of the following vehicle. The cooperative following execution module 220 performs reverse calculations to satisfy the communication distance constraint (i.e., the turning radius is limited to a certain value). And the maximum allowable coupling velocity under dynamic stability boundary conditions. The calculation is based on the following formula: ; Collaborative Follow Execution Module 220 according to Generate braking commands to drive the driving mechanism to decelerate, and simultaneously according to The generated steering command drives the self-balancing mechanism to adjust the roll angle, thereby ensuring that the following vehicle completes the following without leaving the communication range or overturning. The generation of the PWM signal for the motor drive and the underlying PID closed-loop control can be achieved by those skilled in the art using existing motor vector control technology, which is well-known in the field and will not be elaborated upon here.

[0050] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An integrated control system for a self-balancing two-wheeled vehicle with master-slave following capability, characterized in that, include: Master vehicle control terminal, slave vehicle control terminal, and communication module; The main vehicle control terminal includes an intent parsing module and a prediction instruction generation module; The intent parsing module collects the biomechanical signals applied by the rider to the master vehicle and the kinematic data of the master vehicle, parses the driving intent features and uses the kinematic data to perform temporal consistency verification to obtain the verified and valid driving intent features; the prediction instruction generation module maps the verified and valid driving intent features into a predictive control vector containing future time window information, and sends it to the slave vehicle control terminal through the communication module. The slave vehicle control terminal includes an envelope verification module and a cooperative following execution module; the envelope verification module receives the predicted control vector, estimates the physical load parameters of the slave vehicle based on the current suspension travel and load state of the slave vehicle, calculates the dynamic stability boundary, and compares the predicted control vector with the dynamic stability boundary to obtain the verification result; The collaborative follow-up execution module selects a strategy based on the verification result: if the verification result does not exceed the limit, it executes the attitude synchronization strategy; if it exceeds the limit, it executes the trajectory compensation strategy or the velocity-radius coupling adjustment strategy.

2. The integrated control system for a self-balancing two-wheeled vehicle with master-slave following capability as described in claim 1, characterized in that, The step of parsing the driving intention features and performing temporal consistency verification on them using the kinematic data to obtain valid driving intention features includes: The intent parsing module performs spatial synthesis processing on the collected seat pressure distribution data and foot pedal pressure data to calculate the change in the position of the pressure center relative to the geometric center of the main vehicle seat. By combining the collected steering torque input, the pressure center position change and the steering torque input are transformed into a longitudinal intention component representing acceleration and deceleration demand and a lateral intention component representing steering demand using a multi-parameter fusion mapping model, which serve as the driving intention features. Set a verification time window, and calculate the correlation energy between the longitudinal intention component, the lateral intention component and the actual longitudinal acceleration and actual roll rate of the main vehicle within the verification time window; The correlation energy is compared with the confidence threshold. If the correlation energy is greater than the confidence threshold, the driving intention feature is determined to be valid, and the verified valid driving intention feature is obtained.

3. The integrated control system for a self-balancing two-wheeled vehicle with master-slave following capability according to claim 1, characterized in that, The step of the prediction instruction generation module mapping the verified and valid driving intention features into a predictive control vector containing future time window information includes: The longitudinal and lateral intention components in the verified and valid driving intention features are mapped to expected acceleration and expected path curvature using the main vehicle dynamic response model. The target time for the action to take effect is calculated. The target time is obtained by superimposing the current system time, the transmission delay of the communication module, the mechanical response time of the actuator of the vehicle, and the forward prediction time associated with the speed. Determine the communication coverage radius constraint parameters based on the physical characteristics of the communication module; The expected acceleration, the expected path curvature, the target time, and the communication coverage radius constraint parameters are encapsulated into the predictive control vector.

4. The integrated control system for a self-balancing two-wheeled vehicle with master-slave following capability according to claim 1, characterized in that, The steps of the envelope verification module to estimate the physical load parameters of the vehicle and calculate the dynamic stability boundary based on the current suspension travel and load state of the vehicle include: Calculate the current total mass of the vehicle based on the current suspension travel and suspension spring stiffness characteristics of the vehicle. The current center of gravity height of the vehicle is calculated based on the vehicle's unloaded center of gravity height, the preset height of the cargo area, and the vehicle's current total mass. The anti-rollover lateral acceleration threshold is calculated based on the current center of gravity height of the vehicle and the wheel track width. An elliptical envelope region with longitudinal acceleration and lateral acceleration as variables is constructed as the dynamic stability boundary. The boundary of the elliptical envelope region is defined by the longitudinal acceleration limit corresponding to the road friction coefficient and the anti-rollover lateral acceleration threshold.

5. The integrated control system for a self-balancing two-wheeled vehicle with master-slave following capability according to claim 4, characterized in that, The step of the envelope verification module comparing the predicted control vector with the dynamic stability boundary to obtain the verification result includes: The requested lateral acceleration is calculated based on the expected path curvature in the predicted control vector and the vehicle's current speed. Determine whether the state point formed by the expected acceleration in the predicted control vector and the requested lateral acceleration is located within the elliptical envelope region; If the state point is located within the elliptical envelope region and meets the preset safety margin coefficient, then the verification result is determined to be within the limit; Otherwise, the verification result is determined to be out of bounds.

6. The integrated control system for a self-balancing two-wheeled vehicle with master-slave following capability according to claim 1, characterized in that, The steps of the cooperative following execution module in executing the attitude synchronization strategy include: The preset pitch angle to counteract the longitudinal inertial force is calculated based on the expected acceleration in the predicted control vector, and the preset roll angle to counteract the centrifugal force is calculated based on the expected path curvature in the predicted control vector and the current vehicle speed. Generate an interpolated trajectory from the current attitude to the preset pitch angle and the preset roll angle; drive the self-balancing mechanism of the slave vehicle to output a feedforward torque so that the slave vehicle body completes attitude tilting before the target time specified by the predicted control vector; The driving mechanism of the vehicle outputs driving torque according to the expected acceleration.

7. The integrated control system for a self-balancing two-wheeled vehicle with master-slave following capability according to claim 1, characterized in that, When the verification result is out of limit, the cooperative follow-up execution module is specifically used for: The minimum safe turning radius of the vehicle at present is calculated using the anti-rollover lateral acceleration threshold and the safety margin coefficient; The curvature of the planned compensation path corresponds to the minimum safe turning radius and is tangent to the trajectory of the master vehicle; the master-slave distance is predicted after the slave vehicle travels along the compensation path; The master-slave vehicle distance is compared with the communication coverage radius constraint parameter in the predictive control vector. If the master-slave vehicle distance is less than or equal to the communication coverage radius constraint parameter, the trajectory compensation strategy is selected to be executed. If the distance between the master and slave vehicles is greater than the communication coverage radius constraint parameter, then the speed radius coupling adjustment strategy is selected to be executed.

8. The integrated control system for a self-balancing two-wheeled vehicle with master-slave following capability according to claim 7, characterized in that, The steps of the collaborative following execution module in executing the trajectory compensation strategy include: Drive the vehicle's driving mechanism to maintain the current speed; The self-balancing mechanism of the driven vehicle is coordinated with the steering to control the driven vehicle to travel along the compensation path, thereby creating a lateral offset relative to the master vehicle.

9. The integrated control system for a self-balancing two-wheeled vehicle with master-slave following capability according to claim 7, characterized in that, The steps of the cooperative follow-up execution module in executing the speed-radius coupling adjustment strategy include: Based on the communication coverage radius constraint parameters, the anti-rollover lateral acceleration threshold, and the safety margin coefficient, the maximum allowable coupling speed is calculated in reverse. Generate a braking command to drive the vehicle's driving mechanism to decelerate to the maximum permissible coupling speed; The generated steering command drives the self-balancing mechanism of the slave vehicle to adjust the roll angle, controlling the slave vehicle to travel with the turning radius corresponding to the communication coverage radius constraint parameter.

10. The integrated control system for a self-balancing two-wheeled vehicle with master-slave following capability according to claim 1, characterized in that, The communication module is used to establish a bidirectional data link between the master vehicle control terminal and the slave vehicle control terminal using point-to-point wireless technology with ultra-wideband wireless communication protocol or WIFI direct connection.

Images