A Multi-Parameter Collaborative Detection Method for Traction Performance of Mobile Robots

Abstract

This application discloses a multi-parameter collaborative detection method for the traction performance of mobile robots, belonging to the field of mobile robot performance testing technology. By building a collaborative detection system, it achieves synchronous acquisition and accurate calculation of hook traction force, slip ratio, travel speed, and traction power; it employs nanosecond-level timestamps and linear interpolation technology to ensure that the synchronization error of multiple parameters is ≤8μs; it achieves shock-free switching between active drive and passive towing modes through closed-loop control; after preprocessing and algorithm optimization, the relative error of traction force measurement is ≤0.2%, and the relative error of slip ratio measurement is ≤±0.5%. This method solves the problems of lack of testing standards, asynchronous multiple parameters, low testing accuracy, and poor adaptability to working conditions in existing technologies, and can be widely applied to the traction performance evaluation of tracked mobile robots, providing support for product development and quality control.

Description

Technical Field

[0001] This invention relates to the field of mobile robot performance testing technology, specifically to a multi-parameter collaborative testing method for the traction performance of a mobile robot. Background Technology

[0002] Currently, mobile robots, with their excellent ability to navigate complex terrain, have been widely used in key scenarios such as emergency rescue, industrial logistics, and agricultural production. Traction characteristics, as the core indicator for evaluating their power output, load capacity, and environmental adaptability, directly determine the robot's operational efficiency and reliability in tasks such as traction, dragging, climbing, and obstacle crossing.

[0003] Currently, there is a lack of standards for testing the traction performance of mobile robots, and no unified standardized procedures exist in the industry. Existing testing methods mainly rely on simple static tests, which can only roughly estimate the maximum traction force and cannot achieve multi-parameter collaborative evaluation under dynamic working conditions. A simple static test involves fixing one end of a dynamometer to a wall with a rope and connecting the other end to a mobile robot. The robot is then moved away from the dynamometer, and the traction force is visually observed when the robot's wheels slip, which is taken as the robot's traction force. This method cannot simultaneously measure the wheel slip rate and travel speed during robot movement, resulting in insufficient scientific rigor, low measurement accuracy and efficiency. It can only roughly estimate the robot's traction force and cannot accurately capture the traction force value at the point of maximum slip rate, leading to a significant error between the measured traction force value and the actual maximum traction force. Furthermore, it cannot test traction power. Therefore, this simple testing method has significant limitations.

[0004] Therefore, there is an urgent need for a standardized, high-precision, multi-parameter collaborative detection method to solve the problems of insufficient dynamic working condition adaptation, asynchronous parameter measurement, and low test accuracy in existing technologies, and to provide scientific and reliable technical support for the evaluation of the traction characteristics of mobile robots. Summary of the Invention

[0005] This application provides a method for multi-parameter collaborative detection of the traction performance of a mobile robot, which solves the problems in the prior art.

[0006] This application provides a method for multi-parameter collaborative detection of the traction performance of a mobile robot, including the following steps:

[0007] S1. Build a collaborative detection system, which includes the mobile robot under test, an electric-driven full-trailer load vehicle, a multi-source sensor group, a data processing unit, and a remote host computer.

[0008] S2, calibrate each sensor in the multi-source sensor group and set system parameters;

[0009] S3 collects traction force, wheel speed, and driving speed data;

[0010] S4 performs zero-point correction, outlier removal, and noise suppression on the collected traction force, wheel speed, and travel speed data, and performs data preprocessing. Based on the collected traction force, wheel speed, and travel speed data, it calculates the slip ratio and traction power.

[0011] S5, based on the electric-driven full trailer load vehicle and the different ground rolling resistance thresholds, sets up dual-mode fuzzy adaptive PID control and automatic switching control, and judges, selects, switches and controls the active drive mode and passive towing mode of the electric-driven full trailer load vehicle and the traction closed-loop control according to the target traction force value.

[0012] S6 outputs traction performance test reports via a remote host computer.

[0013] Furthermore, the electric-drive full-trailer load vehicle has an electric-drive full-trailer structure, equipped with a drive module and an energy consumption module, and can achieve active drive and passive towing.

[0014] The multi-source sensor group includes a cylindrical resistance strain gauge tension sensor, a wheel speed acquisition sensor, and a dual-antenna RTK inertial navigation system;

[0015] The data processing unit is an embedded host, and a layered modular software system is developed based on the robot operating system;

[0016] The mobile robot under test is detachably connected to the electric-driven trailer load vehicle via a rigid tow bar. A column-type resistance strain gauge tension sensor is connected in series in the middle of the tow bar, a wheel speed acquisition sensor is fixed to the drive wheel axle of the mobile robot under test, and an RTK dual antenna is fixed to the top of the electric-driven trailer load vehicle.

[0017] Furthermore, S2 includes:

[0018] Calibration of the column-type resistance strain gauge tension sensor: Using standard weights of 100kg, 200kg, 500kg, and 1000kg, a linear transformation formula F=K×raw+B is obtained through fitting, where F is the traction force, raw is the digital output of the column-type resistance strain gauge tension sensor, K is the proportional coefficient, and B is the zero-point offset; Magnetic encoder calibration: With the drive wheel statically locked, 1000 sets of angle data are collected, the zero-point offset value is calculated, and subtracted in real time during dynamic testing; RTK calibration: The true trajectory is obtained through a laser tracker; Parameter configuration: Sampling frequency is set to 100Hz for traction force, 50Hz for wheel speed, and 20Hz for travel speed, with a synchronization error threshold of 8μs and a rolling resistance threshold of Frl.

[0019] Furthermore, the slip ratio in S4 is calculated based on the formula S=(v_t - v_a) / v_t ×100%, where v_t is the theoretical speed of the drive wheel, v_t=ω×r, ω is the angular velocity of the drive wheel measured by the wheel speed acquisition sensor, r is the rolling radius of the drive wheel, and v_a is the actual driving speed measured by RTK.

[0020] Traction power calculation: based on the formula P=F×v_a, where P is the traction power, F is the pre-processed traction force, and v_a is the pre-processed driving speed.

[0021] Furthermore, S5 includes:

[0022] The data processing unit compares the measured traction force F with the rolling resistance threshold Frl in real time and automatically switches the control mode.

[0023] In active drive mode, when F < Frl, the electric-driven full-trailer load vehicle maintains a speed lower than that of the mobile robot under test through closed-loop control, generates motor speed commands, and ensures that the traction force accurately tracks the target value. Control indicators: steady-state error ≤ ±4N, response time ≤ 1.9s, speed fluctuation ≤ ±0.02m / s.

[0024] In passive towing mode, when F > Frl, the electric-driven full-trailer load truck stops active driving and outputs braking torque through electromagnetic braking of the motor to generate motor braking command. Control indicators: steady state error ≤ ±5N, response time ≤ 2.3s, speed fluctuation ≤ ±0.03m / s;

[0025] Set a hysteresis threshold Frl±10N to avoid frequent switching of operating conditions; when switching, adjust the torque command through a linear gradual strategy to avoid sudden changes in traction force and control the oscillation time to ≤0.2s.

[0026] Furthermore, the S6 traction performance test report includes traction force, slip ratio, driving speed, and traction power.

[0027] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0028] This invention proposes a multi-parameter collaborative detection method for the traction performance of mobile robots, establishing a comprehensive collaborative detection system for hook traction force, slip ratio, travel speed, and traction power. This overcomes the shortcomings of traditional testing that only focuses on a single traction force parameter, achieving a comprehensive characterization of the traction performance of mobile robots. It achieves high-precision slip ratio testing by combining RTK relative positioning and a magnetic encoder; it employs nanosecond-level timestamps and linear interpolation technology to achieve synchronous acquisition of multiple parameters; and it supports shock-free switching between active drive and passive dragging modes, meeting the full-condition testing requirements of mobile robots from no-load to rated load. Attached Figure Description

[0029] Figure 1 This is a system architecture diagram of the detection method of the present invention;

[0030] Figure 2 This is a block diagram of a dual-mode fuzzy adaptive PID control system.

[0031] Figure 3 This is a flowchart of the data acquisition process for the tension sensor.

[0032] Figure 4 This is a flowchart of the magnetic encoder data acquisition process.

[0033] Figure 5 RTK data acquisition flowchart;

[0034] Figure 6 This is a flowchart of the motor control process. Detailed Implementation

[0035] To facilitate understanding of the present invention, a more complete description of this application will be given below with reference to the accompanying drawings, which illustrate preferred embodiments of the invention. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to enable a more thorough and complete understanding of the disclosure of the present invention.

[0036] It should be noted that the terms "vertical," "horizontal," "up," "down," "left," "right," and similar expressions used in this article are for illustrative purposes only and do not represent the only possible implementation.

[0037] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains; the terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to limit the invention; the term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0038] refer to Figures 1-6 A multi-parameter collaborative detection method for the traction performance of mobile robots includes the following steps:

[0039] A collaborative detection system is established, which includes a mobile robot under test, an electrically driven full-trailer load vehicle, a multi-source sensor group, a data processing unit, and a remote host computer.

[0040] Calibrate each sensor in the multi-source sensor group and set the system parameters;

[0041] The data collected on traction force, wheel speed, and driving speed are linearly interpolated and supplemented based on the unified ROS nanosecond-level timestamp benchmark, with 100Hz as the grid, to achieve multi-parameter time alignment and ensure synchronization accuracy.

[0042] The collected traction force, wheel speed, and travel speed data are preprocessed with zero-point correction, outlier removal, and noise suppression. The slip ratio and traction power are calculated based on the collected traction force, wheel speed, and travel speed data.

[0043] Based on the electric-driven full-trailer load vehicle and the different ground rolling resistance thresholds, a dual-mode fuzzy adaptive PID control and automatic switching control are set up. Based on the target traction force value, the active drive mode and passive towing mode of the electric-driven full-trailer load vehicle are judged, selected, switched, and the traction force is closed-loop controlled.

[0044] The system outputs traction performance test reports via a remote host computer, including: traction force-time curves, slip rate-time curves, driving speed-time curves, traction power-time curves, and key system parameters such as motor status under various operating conditions.

[0045] In this embodiment, the electric-driven full-trailer load vehicle has an electric-driven full-trailer structure, a total weight of 1100kg, a maximum test traction force of 8000N, and is equipped with a drive module and an energy consumption module, supporting both active drive and passive towing working modes.

[0046] The multi-source sensor group includes: a cylindrical resistance strain gauge tension sensor (range 0~10kN, accuracy ≤±0.1% FS), a wheel speed acquisition sensor (resolution ≥12 bits, angle accuracy ≤±0.1°), and a dual-antenna RTK inertial navigation system (positioning accuracy at the centimeter level, heading angle accuracy ≤0.2°).

[0047] The data processing unit adopts an embedded host and develops a layered modular software system based on the Robot Operating System (ROS), which has the functions of synchronous acquisition of multi-source data, preprocessing, control command generation and data transmission.

[0048] The mobile robot under test is detachably connected to the electric-driven full-trailer load vehicle via a rigid tow bar. A column-type resistance strain gauge tension sensor is connected in series in the middle of the tow bar. A wheel speed acquisition sensor is fixed to the drive wheel axle of the mobile robot under test. An RTK dual antenna is fixed to an open area on the top of the electric-driven full-trailer load vehicle to ensure unobstructed signal.

[0049] Using a standard weight set ranging from 0 to 1000 kg, the digital output under different loads is acquired through a signal acquisition module. The linear conversion formula F = K × raw + B (where F is the traction force, raw is the digital output of the sensor, K is the proportional coefficient, and B is the zero-point offset) is obtained by fitting the data. The relative error across the entire range is ≤0.2%.

[0050] The analog signal output by the column-type resistance strain gauge tension sensor is converted into a digital signal by the signal acquisition module and transmitted to the data processing unit through a standardized communication bus.

[0051] The wheel speed acquisition sensor collects the angle data of the drive wheel of the mobile robot under test and transmits it to the data processing unit through wireless communication. The real-time rotation speed of the drive wheel is calculated based on the angle difference-time difference algorithm.

[0052] The dual-antenna RTK inertial navigation system receives satellite signals and differential data from the base station, parses navigation protocol data, and outputs driving speed. When GNSS signals are blocked, the inertial measurement unit (IMU) is activated to perform inertial calculation and compensation, with a compensation time of ≤10s, to ensure continuous driving speed data.

[0053] A unified time reference is established based on nanosecond-level timestamps. The highest sampling frequency is used as the time grid. Linear interpolation is performed on parameter data with different sampling frequencies to achieve time alignment of traction force, wheel speed, and driving speed, with a synchronization error ≤8μs.

[0054] Slip ratio calculation: based on the formula S=(v_t - v_a) / v_t ×100%, where v_t is the theoretical speed of the drive wheel (v_t=ω×r, ω is the angular velocity of the drive wheel measured by the wheel speed acquisition sensor, and r is the rolling radius of the drive wheel), and v_a is the actual driving speed measured by RTK;

[0055] Traction power calculation: based on the formula P=F×v_a, where P is the traction power, F is the pre-processed traction force, and v_a is the pre-processed driving speed;

[0056] The data processing unit compares the measured traction force F with the rolling resistance threshold Frl in real time and automatically switches the control mode.

[0057] In active drive mode (F < Frl), the electric-driven full-trailer load vehicle maintains a speed lower than that of the mobile robot under test through closed-loop control, generates motor speed commands, and ensures that the traction force accurately tracks the target value. Control indicators: steady-state error ≤ ±4N, response time ≤ 1.9s, speed fluctuation ≤ ±0.02m / s.

[0058] In passive towing mode (F > Frl), the electric-driven full-trailer load vehicle stops active driving and outputs braking torque through electromagnetic braking of the motor to generate motor braking command. Control indicators: steady state error ≤ ±5N, response time ≤ 2.3s, speed fluctuation ≤ ±0.03m / s;

[0059] Set a hysteresis threshold Frl±10N to avoid frequent switching of operating conditions; when switching, adjust the torque command through a linear gradual strategy (complete the transition within 20 control cycles) to avoid sudden changes in traction force, with an oscillation time ≤0.2s.

[0060] Example:

[0061] System components:

[0062] The mobile robot under test: RXR-MC80BD tracked firefighting robot (curb weight 520kg, maximum design traction force 3700N, maximum speed 1.5m / s).

[0063] Electric-drive full trailer loader: Electric-drive full trailer structure, with a total weight of 1100kg, equipped with a three-phase permanent magnet synchronous motor (rated power 13kW), drive axle (speed ratio 26.69) and forced air cooling energy consumption module;

[0064] Multi-source sensor group: cylindrical resistance strain gauge tension sensor (range 0~10kN), AS5600 magnetic encoder (12-bit resolution), YIN680 dual-antenna RTK inertial navigation system;

[0065] Data processing unit: Jetson Nano embedded host (quad-core ARM Cortex-A57 processor, 128GB TF card), software system developed based on ROS Melodic;

[0066] Host computer: An industrial computer running LabVIEW 2023, supporting real-time data visualization and command issuance;

[0067] Mechanical connection: The mobile robot under test is connected to the electric-driven full-trailer load vehicle through a rigid traction rod. A column-type resistance strain gauge tension sensor is connected in series in the middle of the traction rod. An AS5600 magnetic encoder is fixed to the drive wheel axle of the mobile robot under test through a magnetic bracket (magnet air gap 1.5mm). An RTK dual antenna is fixed to the top of the electric-driven full-trailer load vehicle (spacing 60cm).

[0068] System calibration:

[0069] Calibration of column-type resistance strain gauge tension sensor: Using standard weights of 100kg, 200kg, 500kg, and 1000kg, the conversion formula F=0.152×raw-2.3 was obtained through fitting, with a relative error of 0.1% across the entire range;

[0070] Magnetic encoder calibration: Drive wheel statically locked, 1000 sets of angle data collected, zero offset value 0.3°, subtracted in real time during dynamic testing;

[0071] RTK calibration: The true trajectory is obtained through a laser tracker, and the speed measurement error under straight-line conditions is 0.0045m / s;

[0072] Parameter configuration: Set sampling frequency (traction force 100Hz, wheel speed 50Hz, driving speed 20Hz), synchronization error threshold 8μs, rolling resistance threshold Frl=215.6N.

[0073] Implementation of the testing process:

[0074] Test conditions: The test surface was a flat, dry cement surface, the wind speed was <0.2m / s, and the RTK satellite signal was locked in a fixed state.

[0075] Pre-test preparation: Start the mobile robot under test and the electric-driven full-trailer load vehicle, pre-drive to a speed of 0.8m / s±0.03m / s, stabilize the traction force at 0N±2N, and maintain the slip rate fluctuation at ±0.2% for 10s before entering the test state;

[0076] Multi-condition testing:

[0077] Active drive mode (50N, 100N, 200N): The host computer issues the target traction force command, and the electrically driven full-trailer load vehicle actively drives the vehicle. The data processing unit synchronously collects and processes the data. Test results show that the steady-state error of traction force is ±3N, ±3N, and ±4N, respectively, the response time is 1.7~1.9s, and the slip rate is 2.5%~4.1%.

[0078] Passive towing mode (500N, 1500N, 3700N): When the electric-driven full-trailer load vehicle switches to electromagnetic braking mode, the steady-state error of traction force is ±3N, ±5N, and ±5N respectively, the response time is 2.1~2.3s, and the slip ratio is 4.9%~6.8%;

[0079] Data storage and output: Test data is stored in both ROS bag and CSV formats, and the host computer generates a test report containing core parameter curves and statistical indicators.

[0080] Test result verification

[0081] Based on actual vehicle testing, the detection results of this method meet the following indicators:

[0082] Traction force measurement: full range relative error ≤0.2%, steady state error ≤±4N in active drive mode and ≤±5N in passive towing mode;

[0083] Slip ratio measurement: relative error ≤ ±0.5%, maximum slip ratio ≤ 6.8%;

[0084] Speed ​​measurement: fluctuation ≤ ±0.03m / s;

[0085] Multi-parameter synchronization: synchronization error ≤ 8μs;

[0086] Mode switching: Oscillation time ≤ 0.2s, no sudden change in traction force.

[0087] The above embodiments demonstrate that the detection method of the present invention has high accuracy, strong adaptability to working conditions, and standardized process, which can effectively solve the technical bottleneck of existing tracked mobile robot traction characteristic detection and provide reliable support for product development and quality control.

[0088] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. For those skilled in the art, the present invention can have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for collaborative detection of multiple parameters of traction performance of a mobile robot, characterized in that, Includes the following steps: S1. Build a collaborative detection system, which includes the mobile robot under test, an electric-driven full-trailer load vehicle, a multi-source sensor group, a data processing unit, and a remote host computer. S2, calibrate each sensor in the multi-source sensor group and set system parameters; S3 collects traction force, wheel speed, and driving speed data; S4 performs zero-point correction, outlier removal, and noise suppression on the collected traction force, wheel speed, and travel speed data, and performs data preprocessing. Based on the collected traction force, wheel speed, and travel speed data, it calculates the slip ratio and traction power. S5, based on the electric-driven full trailer load vehicle and the different ground rolling resistance thresholds, sets up dual-mode fuzzy adaptive PID control and automatic switching control, and judges, selects, switches and controls the active drive mode and passive towing mode of the electric-driven full trailer load vehicle and the traction closed-loop control according to the target traction force value. S6 outputs traction performance test reports via a remote host computer.

2. The method for multi-parameter collaborative detection of traction performance of a mobile robot according to claim 1, characterized in that, In step S1, the electric-driven full-trailer load vehicle has an electric-driven full-trailer structure, equipped with a drive module and an energy consumption module, which can realize active drive and passive towing. The multi-source sensor group includes a cylindrical resistance strain gauge tension sensor, a wheel speed acquisition sensor, and a dual-antenna RTK inertial navigation system; The data processing unit is an embedded host, and a layered modular software system is developed based on the robot operating system; The mobile robot under test is detachably connected to the electric-driven trailer load vehicle via a rigid tow bar. A column-type resistance strain gauge tension sensor is connected in series in the middle of the tow bar, a wheel speed acquisition sensor is fixed to the drive wheel axle of the mobile robot under test, and an RTK dual antenna is fixed to the top of the electric-driven trailer load vehicle.

3. The method for multi-parameter collaborative detection of traction performance of a mobile robot according to claim 1, characterized in that, S2 include: Calibration of the column-type resistance strain gauge tension sensor: Using standard weights of 100kg, 200kg, 500kg, and 1000kg, a linear transformation formula F=K×raw+B is obtained through fitting, where F is the traction force, raw is the digital output of the column-type resistance strain gauge tension sensor, K is the proportional coefficient, and B is the zero-point offset; Magnetic encoder calibration: With the drive wheel statically locked, 1000 sets of angle data are collected, the zero-point offset value is calculated, and subtracted in real time during dynamic testing; RTK calibration: The true trajectory is obtained through a laser tracker; Parameter configuration: Sampling frequency is set to 100Hz for traction force, 50Hz for wheel speed, and 20Hz for travel speed, with a synchronization error threshold of 8μs and a rolling resistance threshold of Frl.

4. The method for multi-parameter collaborative detection of traction performance of a mobile robot according to claim 1, characterized in that, Slip ratio calculation in S4: based on the formula S=(v_t - v_a) / v_t ×100%, where v_t is the theoretical speed of the drive wheel, v_t=ω×r, ω is the angular velocity of the drive wheel measured by the wheel speed acquisition sensor, r is the rolling radius of the drive wheel, and v_a is the actual driving speed measured by RTK; Traction power calculation: based on the formula P=F×v_a, where P is the traction power, F is the pre-processed traction force, and v_a is the pre-processed driving speed.

5. The method for multi-parameter collaborative detection of traction performance of a mobile robot according to claim 1, characterized in that, S5 include: The data processing unit compares the measured traction force F with the rolling resistance threshold Frl in real time and automatically switches the control mode. In active drive mode, when F < Frl, the electric-driven full-trailer load vehicle maintains a speed lower than that of the mobile robot under test through closed-loop control, generates motor speed commands, and ensures that the traction force accurately tracks the target value. Control indicators: steady-state error ≤ ±4N, response time ≤ 1.9s, speed fluctuation ≤ ±0.02m / s. In passive towing mode, when F > Frl, the electric-driven full-trailer load vehicle stops active driving and outputs braking torque through electromagnetic braking of the motor to generate motor braking command. Control indicators: steady state error ≤ ±5N, response time ≤ 2.3s, speed fluctuation ≤ ±0.03m / s.

6. The method for multi-parameter collaborative detection of traction performance of a mobile robot according to claim 1, characterized in that, S5 also includes setting a hysteresis threshold Frl±10N to avoid frequent switching of operating conditions; during switching, the torque command is adjusted through a linear gradual strategy to avoid sudden changes in traction force and control the oscillation time to ≤0.2s.

7. The method for multi-parameter collaborative detection of traction performance of a mobile robot according to claim 1, characterized in that, The S6 traction performance test report includes traction force, slip ratio, driving speed, and traction power.

Images