CN-122018582-A - Robot power and position hybrid control method based on multiple sensors

CN122018582ACN 122018582 ACN122018582 ACN 122018582ACN-122018582-A

Abstract

The invention discloses a multi-sensor-based mobile robot manpower-position hybrid control method, which relates to the technical field of robot control and mainly comprises the steps of obtaining actual rotation angles and positions of wheels through position sensors, obtaining actual running speeds through differential processing, measuring the actual accelerations and angular speeds of a robot through an IMU, mapping the actual speeds into actual force information of a joint space based on a velocity jacobian matrix and a dual relation between the velocity jacobian matrix and a force jacobian matrix, obtaining theoretical accelerations through secondary differentiation of generalized position vectors determined by position data, calculating feedforward compensation force by combining equivalent mass-inertia matrix after comparing the theoretical accelerations with actual accelerations, combining the feedforward compensation force with the actual force information to obtain total compensation force, superposing nominal driving signals generated based on expected paths with the total compensation force to generate a final control command driving motor, and realizing force and position hybrid control of the robot. The invention realizes force observation and compensation without a physical force sensor.

Inventors

ZHONG YING
JIANG HUAN
WANG HAO
Wei Wuzhang
HUANG NING

Assignees

宁波中科奥秘机器人有限公司

Dates

Publication Date: 20260512
Application Date: 20251028

Claims (8)

1. The mobile robot manpower-position hybrid control method based on the multiple sensors is characterized by comprising the following steps of: s1, acquiring an actual rotation angle and an actual position of a wheel of a robot through a position sensor arranged at the wheel of the robot, and differentiating the actual position to obtain an actual running speed of the robot; s2, measuring the actual acceleration and the actual angular velocity of the robot in an operation space in real time through an IMU sensor arranged at the mass center of the robot; S3, based on a velocity jacobian matrix obtained by a robot kinematic model, mapping the actual running velocity into actual force information in a joint space by utilizing a dual relationship of the velocity jacobian matrix and the force jacobian matrix; S4, determining a generalized position vector of the robot based on the actual position and the actual rotation angle, and performing secondary differentiation on the generalized position vector to calculate theoretical acceleration in an ideal state without disturbance; s5, according to the difference between the theoretical acceleration and the actual acceleration, combining an equivalent mass-inertia matrix of the robot, and calculating to obtain feedforward compensation force for compensating inertia disturbance; S6, calculating the total compensation force acting on the robot driving system based on the actual force information and the feedforward compensation force; And S7, superposing a nominal driving signal generated based on the expected path and the total compensation force to generate a final control instruction driving motor.
2. The method for controlling the position mixing of the mobile robot based on the multiple sensors according to claim 1, wherein in the step S5, the equivalent mass-inertia matrix is obtained through an off-line calibration process, and the specific process comprises the steps of driving wheels of the robot to execute preset excitation motions, synchronously recording data of the IMU sensor and the position sensor, and calculating equivalent mass or inertia parameters related to the degrees of freedom respectively through linear regression fit.
3. The multi-sensor-based mobile robot digital hybrid control method according to claim 1, wherein in the step S5, the feedforward compensation force is calculated by calculating a difference between a theoretical acceleration and an actual acceleration, and multiplying the difference by an equivalent mass-inertia matrix to obtain the feedforward compensation force.
4. The multi-sensor-based mobile robot position hybrid control method of claim 1, wherein in step S6, the total compensation force is a sum of a feedforward compensation force and an interference compensation force calculated based on actual force information.
5. The multi-sensor based mobile robot position hybrid control method of claim 4, wherein the disturbance compensation force is calculated by the following formula: In the formula, In order to compensate for the disturbance-compensated force, In the form of a velocity jacobian matrix, For the matrix transposition operation, In order to drive the output torque of the system, As an equivalent mass-inertia matrix, For the actual acceleration to be the case, In the form of a coriolis force matrix, In order to achieve the actual operating speed, Is a gravity matrix.
6. The multi-sensor-based mobile robot position hybrid control method of claim 1, wherein in step S7, the nominal driving signal is generated by calculating a deviation of a desired motion path of the robot from an actual position and an actual running speed via a position controller.
7. The multi-sensor-based mobile robot digital hybrid control method of claim 1, wherein the equivalent mass-inertia matrix is set as a time-varying parameter, and is dynamically updated by an online adaptive algorithm based on long-term statistical differences between theoretical acceleration and actual acceleration.
8. The multi-sensor based mobile robot digital hybrid control method of claim 1, wherein in step S1, the actual running speed is compared with a desired speed to obtain a speed error, and the speed error is used for speed closed loop control to adjust a nominal driving signal.

Description

Robot power and position hybrid control method based on multiple sensors Technical Field The invention relates to the technical field of robot control, in particular to a robot power and position hybrid control method based on multiple sensors. Background In the field of motion control of mobile robots such as Automatic Guided Vehicles (AGVs), the prior art generally adopts wheel speed measurement based on encoders to realize speed closed-loop control, and an Inertial Measurement Unit (IMU) is used for carrying out attitude prediction and inertial compensation. However, this approach relies to a large extent on the ideal kinematic and kinetic model of the robot. In actual operation, uncertainty factors such as ground friction coefficient change, load fluctuation, tire wear and the like can cause significant deviation between an ideal model and an actual system, so that the precision and stability of a traditional control method are reduced. While some of these improvements introduce adaptive algorithms or multi-sensor fusion techniques, such as enhancing environmental awareness with lidar or vision sensors, these solutions have not fundamentally addressed the coupling problem between force control and position control. Particularly, the prior art generally lacks an effective real-time force compensation mechanism for insufficient compensation of inertial disturbance caused by acceleration change, and simultaneously fails to fully excavate and utilize the dual relationship between the velocity jacobian matrix and the force jacobian matrix, so that the system cannot accurately infer and compensate external and internal disturbance according to the motion state under the condition of lacking a direct force sensor. The series of problems enable the motion errors to accumulate with time when the mobile robot runs at a high speed or faces a complex dynamic environment, and the overall control performance, adaptability and energy efficiency are difficult to meet the requirement of high-precision operation. Disclosure of Invention In order to meet the requirement of high-precision operation of a mobile robot, the invention provides a robot power and position hybrid control method based on multiple sensors, which comprises the following steps: s1, acquiring an actual rotation angle and an actual position of a wheel of a robot through a position sensor arranged at the wheel of the robot, and differentiating the actual position to obtain an actual running speed of the robot; s2, measuring the actual acceleration and the actual angular velocity of the robot in an operation space in real time through an IMU sensor arranged at the mass center of the robot; S3, based on a velocity jacobian matrix obtained by a robot kinematic model, mapping the actual running velocity into actual force information in a joint space by utilizing a dual relationship of the velocity jacobian matrix and the force jacobian matrix; S4, determining a generalized position vector of the robot based on the actual position and the actual rotation angle, and performing secondary differentiation on the generalized position vector to calculate theoretical acceleration in an ideal state without disturbance; s5, according to the difference between the theoretical acceleration and the actual acceleration, combining an equivalent mass-inertia matrix of the robot, and calculating to obtain feedforward compensation force for compensating inertia disturbance; S6, calculating the total compensation force acting on the robot driving system based on the actual force information and the feedforward compensation force; And S7, superposing a nominal driving signal generated based on the expected path and the total compensation force to generate a final control instruction driving motor. According to the invention, by fusing the position sensor and IMU data and utilizing the dual relation of the speed and the force jacobian matrix, accurate force observation and compensation are realized without a physical force sensor. The feedforward compensation force is generated through the difference between the theoretical acceleration and the actual acceleration, so that the inertia disturbance and the external disturbance are effectively restrained, the track tracking precision, the motion stability and the energy efficiency of the mobile robot under the complex working condition are obviously improved, and the adaptability to dynamic environments such as ground friction, load change and the like is enhanced. Further, in the step S5, the equivalent mass-inertia matrix is obtained through an off-line calibration process, and the specific process comprises the steps of driving wheels of the robot to execute preset excitation movement, synchronously recording data of the IMU sensor and the position sensor, and calculating equivalent mass or inertia parameters related to the degrees of freedom respectively through linear regression fit. Further, in the step S5, the feedforward c