Truss robot control mechanism, truss robot control method, truss robot electric control system and truss robot electric control method

Abstract

The invention relates to a truss robot control mechanism, a truss robot control method, an electric control system and an electric control method. The control mechanism comprises a driving unit and a control unit, wherein a servo motor is connected with a roller through a planetary reducer to drive the truss to move along the light rail; the synchronous belt assembly comprises two synchronous belts which are parallel to the light rail and are fixed on two sides of the light rail; according to the position detection unit, a magnetic encoder is fixed to a truss end beam through an installation support, and a synchronous belt wheel is arranged on a magnetic encoder shaft and meshed with a synchronous belt. And the control module is used for receiving the position feedback signal of the servo motor and the calibration signal of the magnetic encoder, adjusting the output of the servo motor in real time by comparing the difference between the two signals, and eliminating the slip error of the roller.

Description

Technical Field

[0001] The present invention belongs to the technical field of automatic sandblasting of truss robots, and particularly relates to a control mechanism and a control method for improving the long-distance movement accuracy of the X-axis of a truss robot through a closed-loop control mechanism, which are applicable to the automatic sandblasting operation of large workpieces in a high-dust environment, and a control mechanism and a control method for the long-distance movement closed-loop of a truss robot; a closed-loop electric control system of a truss robot and its control method. Background Art

[0002] In the prior art, the following two schemes are mainly adopted for the X-axis movement of a truss robot:

[0003] ⑴ Servo motor drives roller transmission: The installation is simple, but the rollers are prone to slipping, resulting in the inability to guarantee the movement accuracy.

[0004] ⑵ Rack and pinion transmission: The accuracy is relatively high, but it requires high-precision tracks and complex installation, with high costs and is not suitable for long-distance movement.

[0005] Technical Defects:

[0006] ⑴ For the X-axis movement of a truss robot: The servo motor drives the rollers to move. The installation difficulty is low, but the rollers are prone to slipping during the movement process, resulting in the inability to guarantee the movement accuracy.

[0007] ⑵ For the X-axis movement of a truss robot: The rack is installed on one side of the X-axis guide rail, and the servo motor drives the gear to drive the truss axis to move. The equipment cost is high and the installation difficulty is large, and it is not suitable for long-distance movement.

[0008] ⑶ The accuracy of roller transmission is unreliable due to slipping.

[0009] ⑷ The rack and pinion transmission has high costs, complex installation, and is prone to accuracy decline due to the cumulative error of the rack during long-distance operation.

[0010] ⑸ Roller drive combined with a magnetic encoder: Although the structure is simplified, in a high-dust environment (such as a sandblasting scenario), the magnetic encoder is easily contaminated, resulting in signal distortion and the inability to accurately feedback the position.

[0011] ⑹ Open-loop or single closed-loop control: Only relying on the feedback of the motor encoder, when the rollers slip, the position error cannot be corrected in real time, resulting in insufficient movement accuracy.

[0012] ⑺ Sensor singularity: Only relying on the feedback of the magnetic encoder, when the rollers slip with the track, the system cannot quickly identify the error source, resulting in calibration lag.

[0013] ⑻ Algorithm limitation: The traditional PID algorithm has poor adaptability to non-linear disturbances (such as slipping and load mutation) and insufficient dynamic adjustment ability.

[0014] ⑼ Communication delay: The low-speed communication protocol causes delays in the transmission of sensor data and control instructions, affecting the closed-loop response speed. Summary of the Invention

[0015] The technical problem solved by the present invention is to overcome the deficiencies of the above-mentioned prior art and provide a moving mechanism for the X-axis of a truss robot used for sandblasting large workpieces dozens of meters long in a high-dust sandblasting environment. The servo motor is used to drive the rollers in combination with magnetic encoder calibration to achieve high-precision control of long-distance movement. The closed-loop control can know the real-time position of the movement stably and reliably, solving the situation that the installation of the gear-rack drive for the long-distance movement of the truss is difficult and the cost is expensive due to the need for high-precision tracks; solving the situation that when the traditional motor drives the rollers to move and the rollers slip, the accuracy of the truss cannot be guaranteed; using the servo motor to drive the rollers, the light rail is simple and convenient to install, and the closed-loop control by the magnetic encoder ensures the movement accuracy of the X-axis of the truss, and is used for the control mechanism and control method for the long-distance movement closed-loop of the truss robot. Another technical problem solved by the present invention is to provide a high-precision closed-loop electric control system and its control method realized through multi-sensor fusion, adaptive control algorithm and high-speed communication protocol. Another technical problem solved by the present invention is to provide a closed-loop electric control system for a truss robot, which solves the problems of low control accuracy, poor robustness and response delay in the prior art through a multi-sensor fusion circuit, an adaptive control algorithm and a high-speed communication protocol, and realizes a closed-loop electric control system and its control method for a truss robot with high-precision real-time control under long-distance movement.

[0016] The first technical solution of the present invention is that the control mechanism for the long-distance movement closed-loop of the truss robot is characterized in that it includes:

[0017] Drive unit: The servo motor is connected to the rollers through a planetary reducer to drive the truss to move along the light rail;

[0018] Synchronous belt assembly: Two synchronous belts parallel to the light rail are fixed on both sides of the light rail;

[0019] Position detection unit: Select one of the following:

[0020] ⑴ The magnetic encoder is fixed to the truss end beam through a mounting bracket. A synchronous pulley is provided on the shaft of the magnetic encoder and meshes with the synchronous belt;

[0021] ⑵ The laser reflection strips are laid parallel to the X-axis direction, and the laser distance sensor is installed on the truss end beam for real-time detection of the actual displacement of the truss;

[0022] Control module: Select one of the following:

[0023] ⑴ Receive the position feedback signal of the servo motor and the calibration signal of the magnetic encoder, and adjust the output of the servo motor in real time by comparing the difference between the two to eliminate the roller slip error;

[0024] ⑵ It includes an adaptive PID controller that receives the signal from the servo motor encoder and the position data of the laser range sensor, and compensates for the roller slip error by dynamically adjusting the PID parameters.

[0025] Preferably: the pitch accuracy of the synchronous belt ⑸ is ±0.1 mm, and the meshing clearance with the synchronous pulley ⑼ is ≤0.05 mm; the synchronous belt is a carbon fiber synchronous belt, and the surface of the synchronous pulley is provided with anti-slip lines and meshes with the synchronous belt for transmission. The tension of the synchronous belt is dynamically adjusted by a spring tensioning mechanism.

[0026] Preferably: the pitch accuracy of the synchronous belt is ±0.1 mm, and the meshing clearance with the synchronous pulley is ≤0.05 mm; the synchronous belt is a carbon fiber synchronous belt, and the surface of the synchronous pulley is provided with anti-slip lines and meshes with the synchronous belt for transmission. The tension of the synchronous belt is dynamically adjusted by a spring tensioning mechanism.

[0027] Preferably: the resolution of the magnetic encoder is ≥0.01 mm, and the installation position is coaxial with the rotating shaft of the roller.

[0028] Preferably: the surface of the laser reflection strip is covered with an anti-pollution coating, and bar code marks are arranged at intervals in the reflection area for segmented calibration of the laser range sensor; the adaptive PID controller has a built-in machine learning model that predicts the change of the roller friction coefficient based on historical slip data and adjusts the output torque in real time.

[0029] Preferably: the contact surface between the roller and the light rail is provided with self-cleaning grooves, and piezoelectric ceramic chips are embedded in the grooves to remove the dust on the track through high-frequency vibration; the surface of the roller is covered with a composite material with a high friction coefficient.

[0030] The second technical solution of the present invention is the control method for the closed-loop control mechanism of the long-distance movement of the truss robot, which is characterized in that it includes the following steps:

[0031] ⑴ Start the system: the servo motor drives the roller, and the truss starts to move;

[0032] ⑵ Data acquisition: the servo motor and the magnetic encoder synchronously feedback the position signals;

[0033] ⑶ Error judgment: the controller compares the two signals to detect whether the roller slips;

[0034] ⑷ Closed-loop correction: if there is a deviation, the magnetic encoder data is used as the calibration reference to adjust the motor speed; if there is no deviation, go to the next step;

[0035] ⑸ Continuous monitoring: repeatedly execute steps ⑵ to ⑷ until the truss reaches the target position.

[0036] The third technical solution of the present invention is the closed-loop electric control system of the truss robot, which is characterized in that it includes a multi-sensor fusion module, an adaptive control algorithm, and a high-speed communication module:

[0037] The multi-sensor fusion module integrates magnetic encoders, laser displacement sensors, and IMU data;

[0038] The adaptive control algorithm includes the MPC algorithm and the SMC algorithm; it is a cooperative control framework for MPC and SMC; the MPC algorithm is based on the roller dynamics model, and the prediction time domain is 3 to 5 time steps; the switching condition of the SMC algorithm is that the IMU angular velocity deviation exceeds a preset threshold;

[0039] The high-speed communication module uses the CAN FD bus or Gigabit Ethernet protocol to achieve real-time transmission of sensor data and the controller, with a communication delay ≤ 1 ms; it also includes an integrated hardware accelerator to optimize data packet parsing and priority scheduling.

[0040] The fourth technical solution of the present invention is the closed-loop electric control system of the truss robot, which is characterized in that it includes a signal input module, a controller, an output instruction module, a data fusion and decision-making module, and an actuator:

[0041] The signal input module includes: a servo motor encoder for real-time feedback of the roller position signal; a magnetic encoder or a laser range finder for providing a calibration position signal;

[0042] The controller includes: an error comparison unit for comparing the difference between the servo motor and the calibration signal to detect roller slippage; an adaptive PID control module for dynamically adjusting the PID parameters or switching to the MPC / SMC cooperative control;

[0043] The output instruction module is used to generate a motor speed adjustment instruction and transmit it to the servo driver through the high-speed communication module, and the high-speed communication module selects CAN FD / Gigabit Ethernet;

[0044] The data fusion and decision-making include: a Kalman filter for fusing multi-sensor data, and the multi-sensor data is from the IMU, laser, and magnetic encoder; a machine learning module: for predicting the slippage probability by the LSTM neural network and optimizing the control parameters in real time;

[0045] The signal has a loop path from the actuator → sensor → controller → actuator.

[0046] The fifth technical solution of the present invention is a control method for the closed-loop electric control system of the truss robot based on multi-sensor fusion and adaptive algorithms, which is characterized in that it includes the following steps:

[0047] ⑴ Data acquisition: The magnetic encoder, laser sensor, and IMU synchronously acquire position, velocity, and attitude data.

[0048] ⑵ Data fusion: Based on the Kalman filter algorithm, multi-sensor data is fused to output high-confidence position information.

[0049] ⑶ Control decision-making: The MPC algorithm generates preliminary control instructions; if the IMU angular velocity deviation > threshold is detected, SMC intervention is triggered; the machine learning module adjusts control parameters in real time.

[0050] ⑷ Instruction execution: The instructions are sent to the servo motor driver through the high-speed communication module to drive the roller to move. Preferably: The machine learning module further includes: training the LSTM neural network through historical data to predict the probability of roller slippage and adjusting control parameters in advance: motor torque, reduction ratio; designing an online learning mechanism to update the model weights in real time to adapt to environmental changes.

[0051] Preferably: The machine learning module further includes: training the LSTM neural network through historical data to predict the probability of roller slippage and adjusting control parameters in advance: motor torque, reduction ratio; designing an online learning mechanism to update the model weights in real time to adapt to environmental changes;

[0052] The Kalman filter algorithm in step ⑵ includes introducing sliding mode control as an auxiliary strategy. When an abrupt change in the IMU angular velocity is detected, it switches to the SMC mode to quickly suppress disturbances.

[0053] The MPC algorithm in step ⑷ includes predicting the system state for the next 3 - 5 time steps based on the model predictive control framework and combining the roller dynamics model to dynamically optimize the control input.

[0054] Compared with the prior art, the beneficial effects of the present invention are:

[0055] ⑴ The present invention uses a servo motor to drive the roller, which is simple and convenient to install on the light rail. The closed-loop control by the magnetic encoder ensures the motion accuracy of the truss in the X-axis.

[0056] ⑵ The laser ranging sensor of the present invention is not affected by dust and, combined with anti-pollution reflectors, ensures the stability of position detection and is suitable for high-dust scenarios such as sandblasting.

[0057] ⑶ The adaptive PID controller of the present invention fuses encoder and laser ranging data, corrects the slippage error in real time, and improves the positioning accuracy to ±0.1 mm.

[0058] ⑷ The carbon fiber synchronous belt and spring tensioning mechanism of the present invention reduce the wear of the transmission system, reduce the maintenance frequency, and the comprehensive cost is reduced by 40% compared with the gear rack scheme.

[0059] ⑸ The machine learning model of the present invention optimizes the PID parameters, adapts to different load and speed conditions, improves the system response speed and robustness, and enhances the intelligent control level.

[0060] ⑹ The multi-sensor fusion of the present invention reduces the position error from ±0.5 mm to ±0.1 mm, improving the accuracy.

[0061] ⑺ The robustness of the present invention is enhanced: the adaptive algorithm improves the stability of the system by 40% in scenarios such as slipping and sudden load changes.

[0062] ⑻ The high-speed communication protocol of the present invention shortens the closed-loop response time to within 10 ms, optimizing the real-time performance.

[0063] ⑼ The maintenance cost of the present invention is reduced: the redundant design and the online learning mechanism reduce the downtime caused by sensor failures.

[0064] ⑽ The comparison table of the present invention and the prior art is as follows:

[0065] BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 is a schematic diagram of the overall structure of the control mechanism of the present invention;

[0067] Figure 2 is a top view structural schematic diagram of the control mechanism of the present invention.

[0068] Description of the main component symbols:

[0069] Servo motor ⑴ Planetary reducer ⑵ Light rail ⑶ Timing belt ⑷ Magnetic encoder ⑸ End beam ⑹ Timing belt pulley ⑺ Truss ⑻ DETAILED DESCRIPTION OF THE INVENTION

[0070] The present invention will be further described in detail below with reference to the drawings:

[0071]

[0072] Please refer to Figure 1 、 Figure 2 as shown, the control mechanism for the long-distance motion closed-loop of the truss robot includes:

[0073] Drive unit: The servo motor ⑴ is connected to the planetary reducer ⑵ to drive the moving roller (not shown in the figure) to move along the X-axis light rail ⑶;

[0074] Synchronous belt assembly: Two synchronous belts ⑷ parallel to the light rail ⑶ are fixed on both sides of the light rail ⑶;

[0075] Position detection unit: Select one of the following:

[0076] ⑴ The magnetic encoder ⑸ is fixed to the truss end beam ⑹ through a mounting bracket (not shown in the figure). A synchronous pulley ⑺ is provided on the shaft of the magnetic encoder ⑸ and meshes with the synchronous belt ⑷.

[0077] ⑵ Laser reflection strips (not shown in the figure) are laid parallel in the X-axis direction, and laser distance sensors (not shown in the figure) are installed on the end beam ⑼ of the truss ⑻ for real-time detection of the actual displacement of the truss ⑻.

[0078] Control module: Select one of the following:

[0079] ⑴ Receive the position feedback signal of the servo motor ⑴ and the calibration signal of the magnetic encoder ⑸, and adjust the output of the servo motor ⑴ in real time by comparing the difference between the two to eliminate the slip error of the roller (not shown in the figure).

[0080] ⑵ Include an adaptive PID controller, receive the encoder signal of the servo motor ⑴ and the position data of the laser distance sensor (not shown in the figure), and compensate for the slip error of the roller (not shown in the figure) by dynamically adjusting the PID parameters.

[0081] In this embodiment, the pitch accuracy of the synchronous belt ⑷ is ±0.1 mm, and the meshing clearance with the synchronous pulley ⑺ is ≤0.05 mm; the synchronous belt ⑷ is made of a carbon fiber synchronous belt, and the surface of the synchronous pulley ⑺ is provided with anti-slip lines and meshes with the synchronous belt ⑷ for transmission. The tension of the synchronous belt ⑷ is dynamically adjusted by a spring tensioning mechanism (not shown in the figure).

[0082] In this embodiment, the resolution of the magnetic encoder ⑸ is ≥0.01 mm, and the installation position is coaxial with the rotation axis of the roller (not shown in the figure).

[0083] In this embodiment, the surface of the laser reflection strip (not shown in the figure) is covered with an anti-pollution coating, and bar code marks are arranged at intervals in the reflection area for segmented calibration of the laser distance sensor; the adaptive PID controller has a built-in machine learning model, predicts the change of the roller friction coefficient based on historical slip data, and adjusts the output torque in real time.

[0084] In this embodiment, the contact surface between the roller (not shown in the figure) and the light rail ⑶ is provided with self-cleaning grooves (not shown in the figure), and piezoelectric ceramic chips (not shown in the figure) are embedded in the grooves to remove track dust through high-frequency vibration; the surface of the roller (not shown in the figure) is covered with a high friction coefficient composite material.

[0085] A control method for the closed-loop control mechanism of the long-distance movement of a truss robot includes the following steps:

[0086] ⑴ Start the system: The servo motor ⑴ drives the roller (not shown in the figure), and the truss ⑻ starts to move.

[0087] ⑵ Data acquisition: The servo motor ⑴ and the magnetic encoder ⑸ synchronously feedback position signals.

[0088] ⑶ Error judgment: The controller compares the two signals to detect whether the roller (not shown in the figure) slips.

[0089] ⑷ Closed-loop correction: If there is a deviation, the speed of the motor ⑴ is adjusted with the data of the magnetic encoder ⑸ as the calibration reference; if there is no deviation, proceed to the next step.

[0090] ⑸ Continuous monitoring: Steps ⑵ to ⑷ are executed in a loop until the truss ⑻ reaches the target position.

[0091] The closed-loop electric control system of the truss robot includes a multi-sensor fusion module, an adaptive control algorithm, and a high-speed communication module:

[0092] The multi-sensor fusion module integrates the data of the magnetic encoder, the laser displacement sensor, and the IMU.

[0093] The adaptive control algorithm includes the MPC algorithm and the SMC algorithm; it is a collaborative control framework for MPC and SMC; the MPC algorithm is based on the roller dynamics model, and the prediction time domain is 3 to 5 time steps; the switching condition of the SMC algorithm is that the IMU angular velocity deviation exceeds the preset threshold.

[0094] The high-speed communication module uses the CAN FD bus or the Gigabit Ethernet protocol to achieve real-time transmission of sensor data and the controller, with a communication delay ≤ 1 ms; it also includes an integrated hardware accelerator to optimize packet parsing and priority scheduling.

[0095] Another embodiment of the closed-loop electric control system of the truss robot includes a signal input module, a controller, an output instruction module, a data fusion and decision-making module, and an actuator:

[0096] The signal input module includes: a servo motor encoder for real-time feedback of the roller position signal; a magnetic encoder or a laser range sensor for providing a calibration position signal.

[0097] The controller includes: an error comparison unit for comparing the difference between the servo motor and the calibration signal to detect roller slippage; an adaptive PID control module for dynamically adjusting the PID parameters or switching to the MPC / SMC collaborative control.

[0098] The output instruction module is used to generate a motor speed adjustment instruction and transmit it to the servo driver through the high-speed communication module, and the high-speed communication module selects CAN FD / Gigabit Ethernet.

[0099] The data fusion and decision-making include: a Kalman filter for fusing multi-sensor data, and the multi-sensor data is from the IMU, laser, and magnetic encoder; a machine learning module: used for LSTM neural network to predict the slippage probability and optimize the control parameters in real time.

[0100] The loop path of the signal from the actuator → sensor → controller → actuator.

[0101] A control method for a closed-loop electric control system of a truss robot based on multi-sensor fusion and an adaptive algorithm, comprising the following steps:

[0102] ⑴ Data acquisition: The magnetic encoder, laser sensor, and IMU synchronously acquire position, speed, and attitude data.

[0103] ⑵ Data fusion: Based on the Kalman filter algorithm, fuse multi-sensor data and output high-confidence position information.

[0104] ⑶ Control decision-making: The MPC algorithm generates a preliminary control instruction; if the IMU angular velocity deviation > threshold is detected, trigger SMC intervention; the machine learning module adjusts the control parameters in real time.

[0105] ⑷ Instruction execution: Send the instruction to the servo motor driver through the high-speed communication module to drive the roller to move.

[0106] In this embodiment, the machine learning module further includes: training the LSTM neural network through historical data to predict the probability of roller slippage and adjust the control parameters in advance: motor torque, reduction ratio; designing an online learning mechanism to update the model weights in real time to adapt to environmental changes.

[0107] In this embodiment, the Kalman filter algorithm in the step ⑵ includes introducing sliding mode control as an auxiliary strategy. When an IMU angular velocity mutation is detected, switch to the SMC mode to quickly suppress disturbances; the MPC algorithm in the step ⑷ includes a model predictive control framework, combined with the roller dynamics model, to predict the system state in the next 3 to 5 time steps and dynamically optimize the control input.

[0108] The above are only the preferred embodiments of the present invention. All equivalent changes and modifications made within the scope of the claims of the present invention shall fall within the scope covered by the claims of the present invention.

Claims

1. A control mechanism for the closed loop of long-distance movement of a truss robot, characterized in that, It includes: Drive unit: The servo motor is connected to the roller through a planetary reducer to drive the truss to move along the light rail. Synchronous belt assembly: Two synchronous belts parallel to the light rail are fixed on both sides of the light rail. Position detection unit: Select one of the following: ⑴ The magnetic encoder is fixed to the truss end beam through a mounting bracket. A synchronous belt pulley is provided on the magnetic encoder shaft and meshes with the synchronous belt. ⑵ The laser reflection strips are laid parallel to the X-axis direction. The laser distance sensor is installed on the truss end beam to detect the actual displacement of the truss in real time. Control module: Select one of the following: ⑴ Receive the position feedback signal of the servo motor and the calibration signal of the magnetic encoder, and adjust the output of the servo motor in real time by comparing the difference between the two to eliminate the roller slip error. ⑵ It includes an adaptive PID controller, which receives the encoder signal of the servo motor and the position data of the laser distance sensor, and compensates for the roller slip error by dynamically adjusting the PID parameters.

2. The control mechanism for the long-distance motion closed loop of the truss robot according to claim 1, characterized in that, The pitch accuracy of the synchronous belt is ±0.1 mm, and the meshing clearance with the synchronous belt pulley is ≤0.05 mm; the synchronous belt is made of carbon fiber synchronous belt, and the surface of the synchronous belt pulley is provided with anti-slip lines and meshes with the synchronous belt for transmission. The synchronous belt tension is dynamically adjusted by a spring tensioning mechanism.

3. The control mechanism for the long-distance motion closed loop of the truss robot according to claim 1, characterized in that, The resolution of the magnetic encoder is ≥0.01 mm, and the installation position is coaxial with the rotation axis of the roller.

4. The control mechanism for long-distance motion closed-loop of the truss robot according to claim 1, characterized in that, The surface of the laser reflection strip is covered with an anti-pollution coating, and bar code marks are arranged at intervals in the reflection area for segmented calibration of the laser distance sensor; the adaptive PID controller has a built-in machine learning model, which predicts the change of the roller friction coefficient based on historical slip data and adjusts the output torque in real time.

5. The control mechanism for the long-distance motion closed-loop of the truss robot according to claim 1, wherein, The contact surface between the roller and the light rail is provided with self-cleaning grooves, and piezoelectric ceramic chips are embedded in the grooves to remove track dust through high-frequency vibration; the surface of the roller is covered with a high friction coefficient composite material.

6. A control method for a closed-loop control mechanism of a long-distance motion of a truss robot, characterized in that, It includes the following steps: ⑴ Start the system: The servo motor drives the roller, and the truss starts to move. ⑵ Data acquisition: The servo motor and the magnetic encoder synchronously feedback the position signals. ⑶ Error judgment: The controller compares the two signals to detect whether the roller slips. ⑷ Closed-loop correction: If there is a deviation, take the magnetic encoder data as the calibration reference and adjust the motor speed. If there is no deviation, go to the next step; ⑸ Continuous monitoring: Loop through steps ⑵ to ⑷ until the truss reaches the target position.

7. A closed-loop electric control system for a truss robot, characterized in that, It includes a multi-sensor fusion module, an adaptive control algorithm and a high-speed communication module: The multi-sensor fusion module integrates the magnetic encoder, laser displacement sensor and IMU data. The adaptive control algorithm includes the MPC algorithm and the SMC algorithm; it is a cooperative control framework for MPC and SMC; the MPC algorithm is based on the roller dynamics model, and the prediction time domain is 3 to 5 time steps; the switching condition of the SMC algorithm is that the IMU angular velocity deviation exceeds the preset threshold. The high-speed communication module uses the CAN FD bus or gigabit Ethernet protocol to realize the real-time transmission of sensor data and the controller, and the communication delay is ≤1 ms; it also includes an integrated hardware accelerator to optimize data packet parsing and priority scheduling.

8. A closed-loop electric control system for a truss robot, characterized in that, It includes a signal input module, a controller, an output instruction module, a data fusion and decision-making module and an actuator: The signal input module includes: a servo motor encoder for real-time feedback of the roller position signal; a magnetic encoder or a laser range finder for providing a calibration position signal. The controller includes: an error comparison unit for comparing the difference between the servo motor and the calibration signal to detect roller slippage; an adaptive PID control module for dynamically adjusting the PID parameters or switching to MPC / SMC collaborative control. The output instruction module is used to generate a motor speed adjustment instruction and transmit it to the servo driver through a high-speed communication module, and the high-speed communication module selects CAN FD / gigabit Ethernet. The data fusion and decision-making include: a Kalman filter for fusing multi-sensor data, and the multi-sensor data is sourced from an IMU, a laser, and a magnetic encoder; a machine learning module for predicting the slippage probability using an LSTM neural network and optimizing the control parameters in real time. The signal has a loop path from the actuator → sensor → controller → actuator.

9. A control method for a closed-loop electric control system of a truss robot based on multi-sensor fusion and adaptive algorithm, characterized in that, It includes the following steps: ⑴ Data acquisition: The magnetic encoder, laser sensor, and IMU synchronously acquire position, speed, and attitude data. ⑵ Data fusion: Based on the Kalman filter algorithm, multi-sensor data is fused to output high-confidence position information. ⑶ Control decision-making: The MPC algorithm generates a preliminary control instruction; if an IMU angular velocity deviation > threshold is detected, SMC intervention is triggered. The machine learning module adjusts the control parameters in real time. ⑷ Instruction execution: The instruction is sent to the servo motor driver through the high-speed communication module to drive the roller to move.

10. The control method of the closed-loop electronic control system of the truss robot based on multi-sensor fusion and adaptive algorithm according to claim 9, characterized in that, The machine learning module further includes: training an LSTM neural network with historical data to predict the roller slippage probability and adjusting the control parameters in advance: motor torque, reduction ratio; designing an online learning mechanism to update the model weights in real time to adapt to environmental changes. In the Kalman filter algorithm in step ⑵, a sliding mode control is introduced as an auxiliary strategy, and when an IMU angular velocity mutation is detected, it switches to the SMC mode to quickly suppress disturbances. The MPC algorithm in step ⑷ includes a model predictive control framework, combines the roller dynamics model, predicts the system state for the next 3 to 5 time steps, and dynamically optimizes the control input.

Images