Micro vehicle intelligent driving method and system based on electromagnetic navigation

Abstract

The invention provides a miniature vehicle intelligent driving method and system based on electromagnetic navigation, and the method comprises the steps: sensing the magnetic field change of a pre-laid electromagnetic navigation wire through an inductance coil sensor carried by a miniature vehicle; analyzing the magnetic field change to obtain a left and right inductance voltage difference value, and calculating the offset of the miniature vehicle relative to the target path; based on a multi-algorithm fusion control strategy, generating a target control instruction according to the offset; and performing corresponding control on the miniature vehicle according to the target control instruction. According to the miniature vehicle intelligent driving method and system based on electromagnetic navigation, the problems that a traditional miniature vehicle system is low in navigation precision and poor in environmental adaptability are effectively solved, and the navigation stability of the miniature vehicle system is improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent driving technology, and in particular to a method and system for intelligent driving of miniature vehicles based on electromagnetic navigation. Background Technology

[0002] With the rapid development of autonomous driving technology, traditional real-vehicle research faces challenges such as high cost and risk, while pure software simulation cannot realistically simulate various interference factors in the physical environment. Currently available intelligent vehicle navigation solutions mainly include visual navigation and GPS navigation, which have significant limitations in practical applications: visual navigation is easily affected by changes in lighting conditions, resulting in unstable recognition accuracy; GPS navigation suffers from weak signals and low positioning accuracy indoors or in complex environments, and also incurs high equipment costs. Although some electromagnetic navigation solutions have been proposed, they generally suffer from insufficient path recognition accuracy, poor adaptability to complex road conditions, and low system stability. Furthermore, existing technologies lack a miniaturized intelligent vehicle system that can simultaneously achieve low cost, high accuracy, and strong adaptability, thus failing to provide an effective physical verification platform for intelligent driving algorithms.

[0003] In view of this, there is an urgent need for intelligent driving methods and systems for miniature vehicles based on electromagnetic navigation, in order to at least address the aforementioned shortcomings. Summary of the Invention

[0004] One objective of this invention is to provide an intelligent driving method and system for miniature vehicles based on electromagnetic navigation. The miniature vehicle utilizes a dual-row array of inductor coil sensors to detect changes in the magnetic field. The dual-row layout design expands the detection range and improves resolution, enabling the system to detect path changes in advance and achieve centimeter-level positioning accuracy. The voltage difference between the left and right inductor coils is obtained by analyzing the magnetic field changes, and the difference ratio algorithm is used to calculate the precise offset of the miniature vehicle relative to the target path. A multi-algorithm fusion control strategy based on a dynamic combination of fuzzy control and PID control is employed to generate target control commands based on the offset, reducing steering response time and motor speed fluctuation. Finally, steering control, speed control, and obstacle avoidance control are executed according to the target control commands, effectively solving the problems of low navigation accuracy and poor environmental adaptability in traditional miniature vehicle systems, and improving the stability of the miniature vehicle system's navigation.

[0005] The intelligent driving method for miniature vehicles based on electromagnetic navigation provided in this invention includes: The inductor coil sensor mounted on the miniature vehicle senses changes in the magnetic field of the pre-laid electromagnetic navigation wires. The difference in voltage between the left and right inductors is obtained by analyzing the changes in the magnetic field, and the offset of the miniature vehicle relative to the target path is calculated. Based on a multi-algorithm fusion control strategy, target control commands are generated according to the offset. The miniature vehicle is controlled accordingly based on the target control instructions.

[0006] Preferably, the inductor coil sensor mounted on the miniature vehicle includes a dual-row array inductor coil sensor; Dual-row array inductor coil sensors include: The front row of sensors, with a preset first spacing, is used to detect the existence of a path; The rear row sensors with a preset second spacing are used to calculate the offset. The first spacing is greater than the second spacing.

[0007] Preferably, a sinusoidal alternating current is passed through the electromagnetic navigation wire to form a guiding magnetic field.

[0008] Preferably, the multi-algorithm fusion control strategy includes a dynamic combination of fuzzy control and PID control.

[0009] Preferably, in right-angle curves, angle compensation is performed using gyroscope data.

[0010] Preferably, the miniature vehicle is also equipped with an obstacle detection module and a speed detection module; The obstacle detection module includes photoelectric sensors and ultrasonic arrays, while the speed detection module includes a photoelectric encoder.

[0011] The intelligent driving method for miniature vehicles based on electromagnetic navigation provided in this invention also includes an intelligent driving algorithm verification mechanism: Construct a digital twin environment that matches the physical characteristics of the miniature vehicle; the digital twin environment simulates the electromagnetic navigation system, sensor noise characteristics, and actuator delay; Test intelligent driving algorithms in a digital twin environment to generate simulation data; Deploy intelligent driving algorithms onto a miniature vehicle physical platform and execute the same test cases as in a digital twin environment to collect real vehicle operation data; The deviation characteristics were obtained by comparing the actual vehicle operation data with the simulation data; the deviation characteristics include: trajectory deviation, control quantity deviation, and response delay deviation. Based on the characteristics of the deviation, attribute the deviation to the cause and obtain the deviation attribution results; Based on the deviation attribution results, determine the adjustment plan for the intelligent driving algorithm verification process; After adjusting the intelligent driving algorithm verification process according to the adjustment plan, subsequent verification will be carried out.

[0012] Preferably, deviation attribution is performed based on deviation characteristics to obtain deviation attribution results, including: Based on the deviation characteristics, a multi-dimensional quantitative analysis of the deviation characteristics is performed to obtain a structured deviation characteristic data package. The structured deviation characteristic data package includes trajectory deviation quantification values, control quantity deviation quantification values, and deviation pattern classification labels. Attribution category determination is performed based on the structured deviation feature data package, and the attribution category determination results are obtained; the attribution category determination includes electromagnetic factor determination, control algorithm factor determination, and mechanical execution factor determination; Based on the attribution category determination results, specific factors are identified, and specific factor identification results are obtained. Based on the specific factor identification results and environmental parameter records, output a list of environmental interference factors; The results of specific factor identification and the list of environmental interference factors are combined to form an attribution result list, which is used as the bias attribution result.

[0013] Preferably, the adjustment scheme for the intelligent driving algorithm verification process is determined based on the deviation attribution results, including: Analyze the bias attribution results to determine the first set of attribution factors; Obtain a second set of attribution factors preset for the adjustment direction, which includes hardware adjustment, control parameter adjustment, and algorithm adjustment; Match the first and second attribution factor sets to determine the direction of target adjustment; Input the set of attribution factor values ​​corresponding to the first attribution factor set into the adjustment scheme corresponding to the target adjustment direction to determine the model and obtain the adjustment scheme.

[0014] The miniature vehicle intelligent driving system based on electromagnetic navigation provided in this embodiment of the invention includes: The magnetic field change sensing module is used to sense changes in the magnetic field of the pre-laid electromagnetic navigation wires through an inductor coil sensor mounted on the miniature vehicle. The offset calculation module is used to analyze the changes in the magnetic field to obtain the voltage difference between the left and right inductors, and to calculate the offset of the miniature vehicle relative to the target path. The control command generation module is used to generate target control commands based on the offset according to the multi-algorithm fusion control strategy. The control module is used to control the miniaturized vehicle according to the target control commands. The miniature vehicle intelligent driving system based on electromagnetic navigation also includes an intelligent driving algorithm verification mechanism: Construct a digital twin environment that matches the physical characteristics of the miniature vehicle; the digital twin environment simulates the electromagnetic navigation system, sensor noise characteristics, and actuator delay; Test intelligent driving algorithms in a digital twin environment to generate simulation data; Deploy intelligent driving algorithms onto a miniature vehicle physical platform and execute the same test cases as in a digital twin environment to collect real vehicle operation data; The deviation characteristics were obtained by comparing the actual vehicle operation data with the simulation data; the deviation characteristics include: trajectory deviation, control quantity deviation, and response delay deviation. Based on the characteristics of the deviation, attribute the deviation to the cause and obtain the deviation attribution results; Based on the deviation attribution results, determine the adjustment plan for the intelligent driving algorithm verification process; After adjusting the intelligent driving algorithm verification process according to the adjustment plan, subsequent verification will be carried out. This includes attributing deviations based on their characteristics and obtaining the attribution results, including: Based on the deviation characteristics, a multi-dimensional quantitative analysis of the deviation characteristics is performed to obtain a structured deviation characteristic data package. The structured deviation characteristic data package includes trajectory deviation quantification values, control quantity deviation quantification values, and deviation pattern classification labels. Attribution category determination is performed based on the structured deviation feature data package, and the attribution category determination results are obtained; the attribution category determination includes electromagnetic factor determination, control algorithm factor determination, and mechanical execution factor determination; Based on the attribution category determination results, specific factors are identified, and specific factor identification results are obtained. Based on the specific factor identification results and environmental parameter records, output a list of environmental interference factors; The specific factor identification results and the list of environmental interference factors are combined to form an attribution result list, which is used as the bias attribution result; Among them, the adjustment scheme for the intelligent driving algorithm verification process is determined based on the deviation attribution results, including: Analyze the bias attribution results to determine the first set of attribution factors; Obtain a second set of attribution factors preset for the adjustment direction, which includes hardware adjustment, control parameter adjustment, and algorithm adjustment; Match the first and second attribution factor sets to determine the direction of target adjustment; Input the set of attribution factor values ​​corresponding to the first attribution factor set into the adjustment scheme corresponding to the target adjustment direction to determine the model and obtain the adjustment scheme.

[0015] The beneficial effects of this invention are as follows: This invention relates to a miniature vehicle that utilizes a dual-row array of inductor coil sensors to detect changes in the magnetic field. The dual-row layout expands the detection range and improves resolution, enabling the system to anticipate changes in path direction and achieve centimeter-level positioning accuracy. The voltage difference between the left and right inductor coils is obtained by analyzing the magnetic field changes, and a difference ratio algorithm is used to calculate the precise offset of the miniature vehicle relative to the target path. Based on a multi-algorithm fusion control strategy that dynamically combines fuzzy control and PID control, target control commands are generated according to the offset, reducing steering response time and motor speed fluctuations. Finally, steering control, speed control, and obstacle avoidance control are executed according to the target control commands. This invention effectively solves the problems of low navigation accuracy and poor environmental adaptability in traditional miniature vehicle systems, improving the stability of miniature vehicle navigation.

[0016] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in this application.

[0017] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0018] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of a miniature vehicle intelligent driving method based on electromagnetic navigation in an embodiment of the present invention; Figure 2 This is a schematic diagram of the fuzzy controller structure in an embodiment of the present invention; Figure 3 This is a schematic diagram of the miniaturized vehicle control system mechanism in an embodiment of the present invention; Figure 4 This is a schematic diagram of the vehicle speed control system structure in an embodiment of the present invention; Figure 5 This is a schematic diagram of a miniature vehicle intelligent driving system based on electromagnetic navigation in an embodiment of the present invention. Detailed Implementation

[0019] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0020] This invention provides a method for intelligent driving of miniature vehicles based on electromagnetic navigation, such as... Figure 1 As shown, it includes: Step 1: The inductor coil sensors mounted on the miniature vehicle sense changes in the magnetic field of the pre-laid electromagnetic navigation wires. These inductor coil sensors include a double-row array type. The spacing between the front row sensors of the double-row array type is set to a preset first spacing for detecting path existence; the spacing between the rear row sensors is set to a preset second spacing for calculating offset, with the first spacing being greater than the second spacing. A sinusoidal alternating current is passed through the electromagnetic navigation wires to form a guiding magnetic field. The miniature vehicle also includes an obstacle detection module and a speed detection module. The obstacle detection module includes photoelectric sensors and an ultrasonic array, while the speed detection module includes a photoelectric encoder.

[0021] In this embodiment, the electromagnetic navigation wire refers to the wire that is laid in advance on the center line of the track, with signal generators connected to both ends. A sinusoidal alternating current with a preset frequency (e.g., 25kHz) and current intensity (e.g., 200mA) is passed through it. According to Ampere's circuital law, this current will generate an alternating magnetic field around the wire, that is, a guiding magnetic field.

[0022] In this embodiment, the dual-row array inductor coil sensor consists of two rows of parallel inductor coils: the front row sensors are spaced at a first spacing (e.g., 25mm) for coarse positioning and path presence detection. A larger spacing can achieve a wider detection range and detect changes in path direction in advance; the rear row sensors are spaced at a second spacing (e.g., 15mm) for precise positioning and offset calculation. A smaller spacing provides higher resolution and achieves centimeter-level positioning accuracy.

[0023] Step 2: Analyze the changes in the magnetic field to obtain the voltage difference between the left and right inductors, and calculate the offset of the miniature vehicle relative to the target path.

[0024] In this embodiment, when the miniaturized vehicle is located at the center of the path, the left and right coils are equidistant from the conductor, and the induced voltages are basically equal. When the vehicle body deviates from the center, the induced voltage of the coil closer to the conductor increases, and the induced voltage of the coil farther from the conductor decreases, forming a voltage difference. The difference is obtained through a differential amplifier circuit, and the offset is calculated as follows: offset = (induced voltage of the left coil - induced voltage of the right coil) / (induced voltage of the left coil + induced voltage of the right coil).

[0025] Step 3: Based on the multi-algorithm fusion control strategy, generate target control commands according to the offset; the multi-algorithm fusion control strategy includes a dynamic combination of fuzzy control and PID control, where the fuzzy controller handles nonlinear road conditions and the PID controller provides precise adjustment.

[0026] In this embodiment, fuzzy control refers to simulating the human expert decision-making process based on fuzzy set theory and linguistic rules. Precise inputs (such as offsets and rates of change of offset) are fuzzified into linguistic variables (such as "large positive" and "small negative"), inferred through a preset rule base (such as "IF large offset and increase THEN significant turn"), and then defuzzified back into precise outputs. Figure 2 The fuzzy controller structure shown includes a quantization factor, a fuzzy rule table, and a defuzzification method. It employs triangular or trapezoidal membership functions, and the rule base is established through expert experience or trial and error. When implemented on an STM32 or other MCU, the rule table can be pre-stored as a lookup table to improve computational efficiency.

[0027] In this embodiment, PID control refers to control based on the proportional, integral, and derivative terms of the error (offset), and incremental PID is used.

[0028] In this embodiment, the dynamic combination mechanism includes weight allocation and switching logic for fuzzy control and PID control. Weight allocation refers to dynamically adjusting the weights of fuzzy control and PID control based on the offset magnitude. For example, when |offset| > threshold (e.g., 1cm), fuzzy control takes precedence (70% weight); when |offset| ≤ threshold, PID control takes precedence (70% weight). The switching logic refers to the system automatically selecting the optimal control strategy by judging the track curvature and vehicle speed. For example, in complex road conditions such as right-angle bends, angle compensation is performed using gyroscope data to improve control accuracy.

[0029] Step 4: Control the miniature vehicle accordingly based on the target control instructions.

[0030] In this embodiment, the execution of target control commands includes steering control, speed control, and obstacle avoidance control. The miniature vehicle control system mechanism is as follows: Figure 3 As shown, the structure of the vehicle speed control system is as follows: Figure 4 As shown.

[0031] The working principle and beneficial effects of the above technical solution are as follows: This invention relates to a miniature vehicle that utilizes a dual-row array of inductor coil sensors to detect changes in the magnetic field. The dual-row layout expands the detection range and improves resolution, enabling the system to anticipate changes in path direction and achieve centimeter-level positioning accuracy. The voltage difference between the left and right inductor coils is obtained by analyzing the magnetic field changes, and a difference ratio algorithm is used to calculate the precise offset of the miniature vehicle relative to the target path. Based on a multi-algorithm fusion control strategy that dynamically combines fuzzy control and PID control, target control commands are generated according to the offset, reducing steering response time and motor speed fluctuations. Finally, steering control, speed control, and obstacle avoidance control are executed according to the target control commands. This invention effectively solves the problems of low navigation accuracy and poor environmental adaptability in traditional miniature vehicle systems, improving the stability of miniature vehicle navigation.

[0032] This invention provides a method for intelligent driving of miniature vehicles based on electromagnetic navigation, and also includes an intelligent driving algorithm verification mechanism: Construct a digital twin environment that matches the physical characteristics of the miniature vehicle; the digital twin environment simulates the electromagnetic navigation system, sensor noise characteristics, and actuator delay.

[0033] In this embodiment, the digital twin environment refers to a virtual simulation model that corresponds one-to-one with the physical miniature vehicle system, encompassing mappings across four dimensions: geometry, physics, behavior, and rules. For example, in specific implementation, based on Ampere's circuital law, the magnetic field distribution generated by a 20kHz sinusoidal alternating current in a conductor is simulated, and the electromagnetic field intensity distribution is calculated in real time. A magnetic field model is constructed using finite element analysis software (such as COMSOL). The outputs of real sensors in a non-magnetic field environment are collected, and spectrum analysis is performed to establish an ARMA (Autoregressive Moving Average) noise model. Mechanical delay characteristics are modeled, and the time from receiving a command to completing an action is measured using a step response test to establish a first-order inertial and delay element model.

[0034] Test intelligent driving algorithms in a digital twin environment to generate simulation data.

[0035] In this embodiment, the intelligent driving algorithm is an assisted driving algorithm that needs to be tested. Simulation data is generated: the intelligent driving algorithm is run in a digital twin environment, recording the desired trajectory, control variables, and system response.

[0036] The intelligent driving algorithm is deployed to the miniature vehicle physical platform and the same test cases as the digital twin environment are executed to collect real vehicle operation data.

[0037] In this embodiment, the miniature vehicle physical platform refers to the hardware system corresponding to the miniature vehicle. Test cases identical to the digital twin environment refer to test scenarios used in the digital twin environment to test intelligent driving algorithms. Real vehicle operation data refers to actual operation data collected on the physical platform, including position, speed, control variables, etc., used for comparison with simulation data.

[0038] The deviation characteristics are obtained by comparing the actual vehicle operation data with the simulation data; the deviation characteristics include: trajectory deviation, control quantity deviation, and response delay deviation.

[0039] In this embodiment, trajectory deviation refers to the maximum vertical offset between the actual vehicle trajectory and the simulated trajectory, calculated through the fusion of encoder and electromagnetic sensor data. Control quantity deviation is the difference between the actual PWM duty cycle and the theoretical value of the incremental PID-controlled motor. Response delay deviation refers to the difference between the measured response time and the simulation prediction.

[0040] Based on the deviation characteristics, a multi-dimensional quantitative analysis of the deviation characteristics is performed to obtain a structured deviation feature data package. The structured deviation feature data package includes trajectory deviation quantification values, control quantity deviation quantification values, and deviation pattern classification labels.

[0041] In this embodiment, the structured deviation feature data package is a standardized data structure formed by organizing the extracted deviation features, including: trajectory deviation quantization values ​​(such as lateral offset, longitudinal speed deviation), control quantity deviation quantization values ​​(such as PWM duty cycle deviation, steering angle deviation), and deviation mode classification labels (such as high frequency fluctuations in right-angle bend areas, and cumulative errors in circular areas).

[0042] Attribution category determination is performed based on structured deviation feature data packets, and the attribution category determination results are obtained; the attribution category determination includes electromagnetic factor determination, control algorithm factor determination, and mechanical execution factor determination.

[0043] In this embodiment, electromagnetic factors are determined by analyzing the repeatability of trajectory deviations at specific locations (such as right-angle bends, loops, etc.). If similar trajectory deviations (standard deviation < 0.5 mm) repeatedly occur at the same location and are strongly correlated with the path curvature (correlation coefficient > 0.7), they are determined to be electromagnetic factors. At the same time, the relationship between trajectory deviation and vehicle speed is analyzed. If the trajectory deviation increases significantly with increasing vehicle speed, the confidence level of electromagnetic factors is enhanced.

[0044] In this embodiment, the control algorithm factor is determined by analyzing the temporal relationship between the control quantity deviation and the trajectory deviation. If the control quantity deviation occurs before the trajectory deviation (time difference > 20ms) and the response delay increases significantly (e.g., by 40%) under complex road conditions such as right-angle bends, it is determined to be a control algorithm factor. In this embodiment, the mechanical execution factor is determined by analyzing the repeatability and consistency of the response delay. If the response delay maintains high consistency (coefficient of variation <15%) under various road conditions and the trajectory deviation continues to increase when the control quantity reaches the limit value (e.g., PWM=100%), it is determined to be a mechanical execution factor. Based on the attribution category determination results, specific factors are identified, and specific factor identification results are obtained.

[0045] In this embodiment, the specific factor identification includes: Electromagnetic system factor identification: Verify the effectiveness of "coarse positioning of the front row of dual-row sensors (detecting the existence of the path) and fine positioning of the rear row (calculating the offset)".

[0046] Control system factor identification: Analyze the effect of incremental PID closed-loop regulation of DC motor to suppress inertial influence, and verify the applicability of fuzzy control rules in different curvature road sections.

[0047] Mechanical system factor identification: Measure the tire slip ratio of the miniature vehicle under different ground conditions and test the speed stability of the motor under load changes.

[0048] Based on the specific factor identification results and environmental parameter records, output a list of environmental interference factors.

[0049] In this embodiment, the environmental interference factor list records the impact of the environment on system performance, including: Electromagnetic interference: The influence of surrounding equipment on the induced electromotive force generated by the principle of electromagnetic induction.

[0050] Ground conditions: The effect of different surface materials on wheel grip, and the change in slip ratio is measured.

[0051] The results of specific factor identification and the list of environmental interference factors are combined to form an attribution result list, which is used as the bias attribution result.

[0052] Analyze the bias attribution results to determine the first set of attribution factors.

[0053] In this embodiment, the first attribution factor set consists of specific factor identification results with a confidence level greater than a preset confidence level threshold extracted from the attribution result list and their associated environmental interference factors. For example, the specific factor identification result is that the front row coarse positioning (detecting the existence of the path) and the rear row fine positioning (calculating the offset) of the dual-row sensor layout at position A fails, and the associated environmental interference factor is that there is an interfering device that interferes with the induced electromotive force generated by the electromagnetic induction principle within a preset range at position A.

[0054] Obtain the second set of attribution factors preset for the adjustment direction, which includes hardware adjustment, control parameter adjustment, and algorithm adjustment.

[0055] In this embodiment, the adjustment direction includes hardware adjustment, control parameter adjustment, and algorithm adjustment. The second attribution factor set is a preset adjustment direction mapping table, establishing a correspondence between the attribution factor type set (the second attribution factor set) and the adjustment direction. The preset adjustment direction mapping table is constructed based on historical adjustment records. Hardware adjustment includes adjustments to electromagnetic wire parameters, sensor layout, etc., such as sensor spacing optimization. Control parameter adjustment includes adjustments to PID parameters, fuzzy rule weights, etc. Algorithm adjustment addresses control strategy issues; for example, when handling elements such as right-angle turns, using a gyroscope to measure angles results in greater stability.

[0056] Match the first and second attribution factor sets to determine the direction of target adjustment.

[0057] In this embodiment, the target adjustment direction is the adjustment direction corresponding to the matching of factors in the first attribution factor set and the second attribution factor set.

[0058] Input the set of attribution factor values ​​corresponding to the first attribution factor set into the adjustment scheme corresponding to the target adjustment direction to determine the model and obtain the adjustment scheme.

[0059] In this embodiment, the attribution factor value set corresponding to the first attribution factor set is a specific quantitative description of the first attribution factor set. For example, the probability that the front row sensor at position A failed to detect the path is 78% (other areas <5%), the average offset error calculated by the rear row sensor at position A is 4.3mm, and electromagnetic interference of 22.5kHz (close to the system operating frequency of 20kHz) is detected at position A, with a measured value of 18mG (milligauss), which exceeds the system anti-interference threshold of 10mG.

[0060] In this embodiment, the adjustment scheme determination model is obtained by training a neural network model based on the historical adjustment records of the target adjustment direction mapping table. During training, the input is the attribution factor value set corresponding to the second attribution factor set (the specific quantitative description of the second attribution factor set), and the output is the historical adjustment scheme under the corresponding adjustment direction.

[0061] In this embodiment, the adjustment scheme determines the output of the attribution factor value set corresponding to the first attribution factor set of the model. For example, the operating frequency of the electromagnetic wire in the A location area is adjusted from 20kHz to 25kHz to avoid the 22.5kHz interference source frequency; within a 30cm range before and after the A location, the wire current is increased from 200mA to 280mA to enhance the signal strength and improve the signal-to-noise ratio; the spacing between the front row sensors is optimized from 25mm to 28mm, and the spacing between the rear row sensors is adjusted from 15mm to 18mm to improve the detection sensitivity of the A location area.

[0062] After adjusting the intelligent driving algorithm verification process according to the adjustment plan, subsequent verification will be carried out.

[0063] The working principle and beneficial effects of the above technical solution are as follows: This invention introduces an intelligent driving algorithm verification mechanism. Specifically, it first accurately simulates the electromagnetic navigation system, sensor noise characteristics, and actuator delay in a digital twin environment. After generating simulation data by testing the intelligent driving algorithm in this environment, the same algorithm is deployed to a physical miniature vehicle platform to execute the same test cases and collect real vehicle operation data. By comparing and analyzing the two types of data, characteristics such as trajectory deviation, control quantity deviation, and response delay deviation are obtained, and multi-dimensional quantitative processing is performed to form a structured data package. Based on this data package, hierarchical attribution analysis is performed, starting from three major categories: electromagnetic factors, control algorithm factors, and mechanical execution factors, to determine the attribution category judgment results. After determining the attribution category judgment results, specific factors are identified, and environmental interference factors are identified simultaneously to form an attribution result list.

[0064] The first set of attribution factors with a confidence level greater than a preset confidence threshold is extracted from the attribution results list. This first set of attribution factors is then matched with the second set of attribution factors in an adjustment direction mapping table constructed based on historical adjustment records to determine the target adjustment direction. After determining the target adjustment direction, the adjustment scheme determination model, trained using the historical adjustment records corresponding to that direction, determines the final adjustment scheme. This hierarchical attribution matching and detailed determination of the adjustment scheme improves attribution efficiency. Subsequent validation is performed after the adjustment is implemented, forming a complete closed loop.

[0065] This invention significantly improves algorithm verification efficiency and enhances the robustness of the verification platform in complex environments through the synergistic iteration of digital twins and physical verification, providing an efficient verification path for intelligent driving algorithms from the laboratory to practical applications.

[0066] This invention provides a miniature vehicle intelligent driving system based on electromagnetic navigation, such as... Figure 5 The following are examples: The magnetic field change sensing module 1 is used to sense the magnetic field changes of the pre-laid electromagnetic navigation wires through the inductor coil sensor mounted on the miniature vehicle. Offset calculation module 2 is used to analyze the magnetic field changes to obtain the voltage difference between the left and right inductors, and to calculate the offset of the miniature vehicle relative to the target path. Control command generation module 3 is used to generate target control commands based on the offset according to the multi-algorithm fusion control strategy. Control module 4 is used to control the miniaturized vehicle according to the target control instructions; The miniature vehicle intelligent driving system based on electromagnetic navigation also includes an intelligent driving algorithm verification mechanism: Construct a digital twin environment that matches the physical characteristics of the miniature vehicle; the digital twin environment simulates the electromagnetic navigation system, sensor noise characteristics, and actuator delay; Test intelligent driving algorithms in a digital twin environment to generate simulation data; Deploy intelligent driving algorithms onto a miniature vehicle physical platform and execute the same test cases as in a digital twin environment to collect real vehicle operation data; The deviation characteristics were obtained by comparing the actual vehicle operation data with the simulation data; the deviation characteristics include: trajectory deviation, control quantity deviation, and response delay deviation. Based on the characteristics of the deviation, attribute the deviation to the cause and obtain the deviation attribution results; Based on the deviation attribution results, determine the adjustment plan for the intelligent driving algorithm verification process; After adjusting the intelligent driving algorithm verification process according to the adjustment plan, subsequent verification will be carried out. This includes attributing deviations based on their characteristics and obtaining the attribution results, including: Based on the deviation characteristics, a multi-dimensional quantitative analysis of the deviation characteristics is performed to obtain a structured deviation characteristic data package. The structured deviation characteristic data package includes trajectory deviation quantification values, control quantity deviation quantification values, and deviation pattern classification labels. Attribution category determination is performed based on the structured deviation feature data package, and the attribution category determination results are obtained; the attribution category determination includes electromagnetic factor determination, control algorithm factor determination, and mechanical execution factor determination; Based on the attribution category determination results, specific factors are identified, and specific factor identification results are obtained. Based on the specific factor identification results and environmental parameter records, output a list of environmental interference factors; The specific factor identification results and the list of environmental interference factors are combined to form an attribution result list, which is used as the bias attribution result; Among them, the adjustment scheme for the intelligent driving algorithm verification process is determined based on the deviation attribution results, including: Analyze the bias attribution results to determine the first set of attribution factors; Obtain a second set of attribution factors preset for the adjustment direction, which includes hardware adjustment, control parameter adjustment, and algorithm adjustment; Match the first and second attribution factor sets to determine the direction of target adjustment; Input the set of attribution factor values ​​corresponding to the first attribution factor set into the adjustment scheme corresponding to the target adjustment direction to determine the model and obtain the adjustment scheme.

[0067] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for intelligent driving of a microcar based on electromagnetic navigation, characterized in that, The method comprises the following steps: An inductive coil sensor carried by the micro vehicle senses the magnetic field change of the pre-laid electromagnetic navigation wire; The magnetic field change is analyzed to obtain the left and right inductive voltage difference, and the offset of the micro vehicle relative to the target path is calculated; Based on the multi-algorithm fusion control strategy, the target control instruction is generated according to the offset; According to the target control instruction, the corresponding control of the micro vehicle is carried out.

2. The electromagnetic navigation-based microcar intelligent driving method according to claim 1, wherein, The inductive coil sensor carried by the micro vehicle comprises a double-row array inductive coil sensor; The double-row array inductive coil sensor comprises: The front row sensor with a preset first spacing is used to detect the existence of the path; The rear row sensor with a preset second spacing is used to calculate the offset; Wherein, the first spacing is greater than the second spacing. 3.The electromagnetic navigation-based microcar intelligent driving method of claim 1, wherein, The electromagnetic navigation wire is passed through a sinusoidal alternating current to form a guide magnetic field. 4.The electromagnetic navigation-based microcar intelligent driving method of claim 1, wherein, The multi-algorithm fusion control strategy comprises a dynamic combination of fuzzy control and PID control. 5.The electromagnetic navigation-based microcar intelligent driving method of claim 1, wherein, In the case of a right-angle bend, the gyro data is combined for angle compensation. 6.The electromagnetic navigation-based microcar intelligent driving method of claim 1, wherein, The micro vehicle also carries an obstacle detection module and a speed detection module; Wherein, the obstacle detection module comprises a photoelectric sensor and an ultrasonic array, and the speed detection module comprises a photoelectric encoder.

7. The electromagnetic navigation-based microcar intelligent driving method according to claim 1, wherein, It also includes an intelligent driving algorithm verification mechanism: A digital twin environment consistent with the physical characteristics of the micro vehicle is constructed; the digital twin environment simulates the electromagnetic navigation system, sensor noise characteristics, and actuator delay; The intelligent driving algorithm is tested in the digital twin environment to generate simulation data; The intelligent driving algorithm is deployed to the physical platform of the micro vehicle and executes the same test cases as the digital twin environment to collect real vehicle operation data; The deviation characteristics are obtained by comparing the real vehicle operation data and the simulation data; The deviation characteristics include: trajectory deviation, control amount deviation, and response time delay deviation; According to the deviation characteristics, the deviation attribution result is obtained; According to the deviation attribution result, the adjustment scheme of the intelligent driving algorithm verification process is determined; According to the adjustment scheme, the intelligent driving algorithm verification process is adjusted, and then subsequent verification is carried out. 8.The electromagnetic navigation-based microcar intelligent driving method of claim 7, wherein, According to the deviation characteristics, the deviation attribution result is obtained, including: According to the deviation characteristics, the multi-dimensional quantitative analysis of the deviation characteristics is carried out to obtain a structured deviation characteristic data package, which contains trajectory deviation quantitative value, control amount deviation quantitative value, and deviation mode classification label; According to the structured deviation characteristic data package, the attribution major category judgment result is obtained; the attribution major category judgment includes electromagnetic factor judgment, control algorithm factor judgment, and mechanical execution factor judgment; According to the attribution major category judgment result, the specific factor identification result is obtained; According to the specific factor identification result and the environment parameter record, an environment interference factor list is outputted; The specific factor identification result and the environment interference factor list are combined to form an attribution result list, which is taken as the deviation attribution result. 9.The electromagnetic navigation-based microcar intelligent driving method of claim 7, wherein, According to the deviation attribution result, the adjustment scheme of the intelligent driving algorithm verification process is determined, including: Analyzing the deviation attribution result to determine a first attribution factor set; A second attribution factor set with a preset adjustment direction is obtained, and the adjustment direction includes hardware adjustment, control parameter adjustment, and algorithm adjustment; The first attribution factor set and the second attribution factor set are matched to determine a target adjustment direction; The attribution factor value set corresponding to the first attribution factor set is input into an adjustment scheme determination model corresponding to the target adjustment direction to obtain an adjustment scheme.

10. An intelligent driving system for a miniature car based on electromagnetic navigation, characterized in that, Comprise: A magnetic field change sensing module for sensing the magnetic field change of the pre-laid electromagnetic navigation wire through the inductive coil sensor carried by the micro vehicle; An offset calculation module for analyzing the magnetic field change to obtain the left and right inductive voltage difference and calculating the offset of the micro vehicle relative to the target path; A control instruction generation module for generating a target control instruction based on a multi-algorithm fusion control strategy according to the offset; A control module for controlling the micro vehicle according to the target control instruction.

Images