Tracked robot body intelligent control system based on microkernel controller

Abstract

A tracked robot body intelligent control system based on a microkernel controller comprises a visual image module, a gas detection module, an ultrasonic radar, an IMU (inertial measurement unit), a battery management module, a motor control module, a man-machine interaction module and the microkernel controller. The method can obtain motor control parameters and further control real-time driving actions of the tracked robot, the control process is accurate, efficient and low in delay, the control response is 60-80 microseconds, the microkernel system occupies few resources, the decision-making adaptability is high, and the method can adapt to automatic operation under dynamic environment changes such as obstacle movement, gradient increase and abnormal temperature. And in the low-power mode, the maximum power is limited through automatic starting, the service life of the battery is prolonged, and the problems of control response lag, poor environmental adaptability, interaction decision stiffness and the like of an existing tracked robot control technology in a complex unstructured environment can be effectively solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of tracked robot body intelligence control, and in particular to a tracked robot body intelligence control system based on a microkernel controller. BACKGROUND

[0002] The tracked robot control technology has problems of control response lag, poor environmental adaptability, rigid interactive decision and difficult secondary development in a complex unstructured environment. With the rapid expansion of robot application scenarios from a single structured environment to an unstructured complex environment, the traditional "perception-decision-execution" layered control architecture has been difficult to meet the task requirements of tracked robots in high mobility, high real-time and high safety scenarios. Existing tracked robots mostly use an integrated controller with ARM Cortex-M / R or x86 as the core, and all driving, sensing, decision-making and communication modules are compiled into a single image. When a function needs to be upgraded, such as adding a SLAM algorithm or replacing a laser radar model, the whole system must be recompiled and reprogrammed, resulting in high maintenance cost and serious version fragmentation. Moreover, if the control unit still uses a bare machine MCU as the core, it can only meet the millisecond-level motor closed loop and cannot load large models for scene understanding and task decision-making; if a GPU / NPU is externally connected, the communication delay is >5 ms, resulting in "perception-control" closed loop jitter and unable to meet the requirements of off-road high-speed mobility.

[0003] In view of the above problems, there have been attempts to adopt a new paradigm of "microkernel + body intelligence", such as using a microkernel architecture to achieve high real-time determinacy, distributed collaborative computing and security isolation. However, the existing technology has not solved the deep adaptation problem between the microkernel and the high-torque motor drive, suspension-track coupled dynamics and complex terrain adaptive control specific to the tracked chassis; there is also a lack of special acceleration and resource scheduling mechanism for body intelligence algorithms, which cannot meet the multiple requirements of real-time, scalability, security and energy consumption. SUMMARY

[0004] To solve the problems of control response lag, poor environmental adaptability and rigid interactive decision of the existing tracked robot control technology in a complex unstructured environment, the present application provides a tracked robot body intelligence control system based on a microkernel controller.

[0005] The technical solution adopted by the present application to solve the above technical problems is: a tracked robot body intelligence control system based on a microkernel controller, comprising a visual image module, a gas detection module, an ultrasonic radar, an IMU inertial measurement unit, a battery management module, a motor control module, a human-computer interaction module and a microkernel controller. The visual image module comprises three cameras installed on the front and rear sides of the tracked robot respectively, and can capture images of the surrounding environment of the tracked robot. The gas detection module can collect and detect harmful gases from the surrounding environment of the tracked robot. The ultrasonic radar can measure the distance between the obstacle and the tracked robot. The IMU inertial measurement unit can detect the body posture information of the tracked robot, including linear acceleration and angular acceleration in a three-dimensional coordinate system, and obtain the three-dimensional velocity, displacement, pitch attitude, yaw attitude and roll attitude of the tracked robot through integration and filtering. The battery management module can monitor the battery information in real time, and alarm when the power is lower than the set threshold. The motor control module is used to receive the control instructions sent by the microkernel controller, and control the tracked robot to switch to different driving modes through a differential control model, including straight driving, turning driving, in-place turning, step driving and climbing driving. The human-computer interaction module comprises a remote control terminal and a visual interface, which can issue instructions to the microkernel controller in real time, and display the system information of the microkernel controller. The microkernel controller can receive the data detected and transmitted by the visual image module, temperature detection module, gas detection module, ultrasonic radar, IMU inertial measurement unit and battery management module, and determine the real-time driving posture of the tracked robot according to the body posture information. In combination with the positions of harmful gases and obstacles in the surrounding environment of the tracked robot, the expected control parameters of the motor control module are obtained. The microkernel controller can also obtain the maximum allowable operating power of the tracked robot according to the power and temperature information of the battery, and correct the expected control parameters according to the maximum allowable operating power to obtain the actual control parameters of the motor control module, and convert the actual control parameters into control instructions and issue them to the motor control module.

[0006] Preferably, the communication interfaces between different modules adopt any one or more of SPI, I2C, CAN, RS232, RS485 and USB, and the parameters of the communication interfaces are configured in advance before the system runs.

[0007] Preferably, the parameters of the communication interfaces include port number, device ID, communication rate, protocol type, data type, data length and verification method.

[0008] Preferably, the tracked robot is provided with a lifting gimbal, and a water cannon is installed on the lifting gimbal. The body intelligent control system further comprises a gimbal control module. The microkernel controller can issue control instructions to the gimbal control module. The gimbal control module can control the horizontal, pitch and lifting of the lifting gimbal, and control the opening degree of the electromagnetic valve of the water cannon.

[0009] Preferably, the microkernel controller receives the fault monitoring information of the motor control module in real time and performs closed-loop feedback control.

[0010] Preferably, the microkernel controller adopts an embedded microkernel system, has a modular architecture of task scheduling, body decision-making, data interaction and file system, and the data transmission delay is 60-80 microseconds.

[0011] According to the above technical solution, the application has the following advantages: The application sets multiple different types of sensors on the tracked robot, and changes the control core device to a microkernel controller. After the microkernel controller receives the detection data of the sensors, the motor control parameters can be obtained and the real-time driving action of the tracked robot can be controlled. The control process is accurate, efficient and low in delay, the control response is 60-80 microseconds, the microkernel system occupies less resources, the decision-making adaptability is strong, and the automatic operation can be adapted to dynamic changes of the environment such as obstacle movement, slope increase and temperature anomaly. In low power mode, the maximum power is automatically started to improve the battery life. The application can effectively solve the problems of existing tracked robot control technology in complex unstructured environment, such as control response lag, poor environmental adaptability and rigid interactive decision-making. The modular architecture of the microkernel controller reduces the difficulty of secondary development. DETAILED DESCRIPTION

[0012] The embodiment provides a tracked robot body intelligent control system based on a microkernel controller, which comprises a visual image module, a gas detection module, an ultrasonic radar, an IMU inertial measurement unit, a battery management module, a motor control module, a human-computer interaction module and a microkernel controller.

[0013] The visual image module comprises three cameras installed on the front and rear sides of the tracked robot respectively, and can capture the surrounding environment image of the tracked robot.

[0014] The gas detection module can collect and detect harmful gases from the surrounding environment of the tracked robot.

[0015] The ultrasonic radar can measure the distance between the obstacle and the tracked robot.

[0016] The IMU inertial measurement unit can detect the body posture information of the tracked robot, including linear acceleration and angular acceleration in a three-dimensional coordinate system, and obtain the three-dimensional velocity, mileage displacement, pitch attitude, yaw attitude and roll attitude of the tracked robot through integration and filtering.

[0017] The battery management module can monitor the battery information in real time and alarm when the power is lower than the set threshold.

[0018] The motor control module is used for receiving the control instruction issued by the microkernel controller, and controlling the tracked robot to switch to different driving modes, including straight driving, turning driving, in-place turning, step driving and climbing driving, through a differential control model.

[0019] The human-computer interaction module includes a remote control terminal and a visualization interface, and can issue instructions to the microkernel controller in real time and display system information of the microkernel controller.

[0020] The microkernel controller can receive data detected and transmitted by the visual image module, the temperature detection module, the gas detection module, the ultrasonic radar, the IMU inertial measurement unit and the battery management module, and can determine the real-time driving posture of the tracked robot according to the somatic posture information, and obtain expected control parameters of the motor control module in combination with the positions of harmful gases and obstacles in the surrounding environment of the tracked robot. Specifically, when the obstacle distance is < 70 cm, the tracked robot retreats or waits in place; when the obstacle distance is 70-80 cm, the tracked robot turns; when the obstacle distance is > 80 cm, the tracked robot continues to drive straight; when the pitch angle is > 10° and the acceleration is > 5 m / s, the tracked robot switches to the step driving mode; and when the pitch angle is > 10° and the acceleration is ≤ 5 m / s, the tracked robot switches to the climbing mode.

[0021] The microkernel controller can also obtain the maximum allowable operating power of the tracked robot according to the battery power and temperature information, and when the power is < 40%, the power saving mode is started, the system power is reduced to 60%, and the expected control parameters are corrected according to the maximum allowable operating power to obtain actual control parameters of the motor control module, and the actual control parameters are converted into control instructions and issued to the motor control module.

[0022] The communication interfaces between different modules adopt any one or more of SPI, I2C, CAN, RS232, RS485 and USB, and the parameters of the communication interfaces are configured in advance before the system is operated. The parameters of the communication interfaces include port number, device ID, communication rate, protocol type, data type, data length and verification method.

[0023] Some types of tracked robots are provided with a lifting gimbal, and a water cannon is installed on the lifting gimbal, such as a fire rescue robot. For this type of tracked robot, the somatic intelligent control system further includes a gimbal control module, the microkernel controller can issue control instructions to the gimbal control module, and the gimbal control module can control the horizontal, pitch and lifting of the lifting gimbal, and control the opening degree of the electromagnetic valve of the water cannon.

[0024] The microkernel controller adopts an embedded microkernel system, has a modular architecture of task scheduling, embodiment decision, data interaction and file system, and a data transmission delay of 60-80 microseconds. The microkernel controller can receive fault monitoring information of the motor control module in real time and perform closed-loop feedback control.

[0025] The control system of the embodiment performs specific control operation as follows: First, the embodiment intelligent control system environment configuration is performed: according to the communication interface of the sensor and the control unit, the communication interface driver (such as SPI, I2C, CAN, RS232, RS485, USB, network interface, etc.) is adapted, and the parameters of the communication interface (such as port number, device ID, communication rate, protocol type, data type, data length, verification method, data packet definition, etc.) is configured.

[0026] Then, the control system data and sensor data processing is performed: multi-channel sensor information such as ultrasonic wave, camera, temperature and tracked robot embodiment state information such as IMU, motor, BMS is collected, the collected data is read and stored in the memory, and the microkernel controller automatically calculates the data according to the protocol after receiving the control instruction, adopts MOSBUS, CAN and free protocol, adopts multiple verification methods to ensure the accuracy and synchronization of the data, and the microkernel multi-process architecture also provides multiple channels for the system to process quickly.

[0027] Next, the embodiment intelligent control system decision is performed: according to the data of the ultrasonic radar and the IMU attitude, the expected control parameters of the motor are matched, it is judged whether the tracked robot is straight turning or obstacle avoidance, flat ground climbing or steps, and the maximum power allowed by the motor current state parameters and BMS is adjusted, and the tracked robot fusion motion state is obtained.

[0028] Finally, the control instruction is executed: the microkernel controller converts the control parameters of the system decision into control instructions through the CAN bus and issues them to the motor driver, and the tracked chassis motor driver and water cannon control unit perform corresponding actions according to the driving instructions.

[0029] In addition, real-time state monitoring and fault handling are also performed: through real-time closed-loop monitoring of the motor control system, the current, voltage, position, temperature, speed and other information of the motor can be detected online, the fault code can be actively reported, and closed-loop feedback control can be performed.

Claims

1. A microkernel-based intelligent control system for a tracked robot, characterized in that: The intelligent control system comprises a visual image module, a gas detection module, an ultrasonic radar, an IMU inertial measurement unit, a battery management module, a motor control module, a human-computer interaction module and a microkernel controller. The visual image module comprises three cameras installed on the front and rear sides of the tracked robot respectively, which can capture images of the surrounding environment of the tracked robot. The gas detection module can collect and detect harmful gases from the surrounding environment of the tracked robot. The ultrasonic radar can measure the distance between the tracked robot and the obstacles. The IMU inertial measurement unit can detect the body posture information of the tracked robot, including linear acceleration and angular acceleration in a three-dimensional coordinate system, and obtain the three-dimensional velocity, displacement, pitch attitude, yaw attitude and roll attitude of the tracked robot through integration and filtering. The battery management module can monitor the battery information in real time and alarm when the battery level is lower than the set threshold. The motor control module is used to receive the control instructions sent by the microkernel controller, control the tracked robot to switch to different driving modes through a differential control model, including straight driving, turning driving, in-place turning, step driving and climbing driving. The human-computer interaction module comprises a remote control terminal and a visual interface, which can send instructions to the microkernel controller in real time and display the system information of the microkernel controller. The microkernel controller can receive the data detected and transmitted by the visual image module, the temperature detection module, the gas detection module, the ultrasonic radar, the IMU inertial measurement unit and the battery management module, determine the real-time driving posture of the tracked robot according to the body posture information, obtain the expected control parameters of the motor control module in combination with the positions of harmful gases and obstacles in the surrounding environment of the tracked robot, correct the expected control parameters according to the maximum allowable operating power of the tracked robot obtained according to the battery level and temperature information, obtain the actual control parameters of the motor control module, and convert the actual control parameters into control instructions and send them to the motor control module.

2. The microkernel controller-based tracked robot body intelligence control system according to claim 1, characterized in that: The communication interfaces between different modules adopt any one or more of SPI, I2C, CAN, RS232, RS485 and USB, and the parameters of the communication interfaces are configured in advance before the system runs.

3. The microkernel controller based tracked robot body intelligence control system of claim 2, wherein: The parameters of the communication interfaces include port number, device ID, communication rate, protocol type, data type, data length and verification method.

4. The microkernel controller based tracked robot body intelligence control system of claim 1, wherein: The tracked robot is provided with a lifting gimbal, and a water cannon is installed on the lifting gimbal.

5. The microkernel controller based tracked robot body intelligence control system of claim 1, wherein: The microkernel controller can send control instructions to the gimbal control module, and the gimbal control module can control the horizontal, pitch and lifting of the lifting gimbal and control the opening of the electromagnetic valve of the water cannon.

6. The microkernel controller based tracked robot body intelligence control system of claim 1, wherein: The microkernel controller receives the fault monitoring information of the motor control module in real time and performs closed-loop feedback control. The microkernel controller adopts an embedded microkernel system, has a modular architecture of task scheduling, body decision, data interaction and file system, and the data transmission delay is 60-80 microseconds.