Mechanical arm tail end contact force soft measurement system based on robot operating system

Abstract

The invention discloses a mechanical arm tail end contact force soft measurement system based on a robot operating system. The system is composed of a sensor array, a PLC, ROS related nodes and a Qt human-computer interaction interface. The PLC preprocesses various signals output by the sensor, uniformly adjusts the signals into a signal format of CAN communication and transmits the signals to the upper computer through a CAN communication protocol, related ROS nodes in the upper computer process the transmitted data and complete conversion of message types, then the data are transmitted to the contact force resolving node to calculate the tail end contact force, and the tail end contact force is calculated. And finally, various data and calculation results are sent to a Qt human-computer interaction interface. According to the measuring system, sensors are installed at all joints and a hydraulic cylinder of the hydraulic excavator, sensor information is preprocessed through a PLC and sent to an ROS upper computer, and finally the upper computer calculates and displays the real-time contact force of the excavator.

Description

Technical Field

[0001] The present invention relates to the fields of embedded systems, sensor technology, and computer science, and particularly to a soft measurement system for the contact force at the end of a robotic arm based on the Robot Operating System (ROS). Background Art

[0002] In recent years, there has been a huge demand for various construction machinery. As a representative of construction machinery, excavators are indispensable in actual operations. However, the traditional operation mode of manually operating excavators not only has low excavation efficiency, and operators must undergo long-term training to complete high-quality operations, but also has a harsh operation environment, posing a threat to personal safety. Therefore, improving the degree of mechatronics of excavators and thus realizing the automatic control of excavators is undoubtedly the most ideal solution to the above problems. Automated excavators have the advantages of low environmental requirements and high excavation efficiency. Universities and research institutions at home and abroad have spared no effort in realizing the automation of excavators and improving control accuracy.

[0003] One of the key issues in the automatic control of excavators is the automatic control of the robotic arm. However, high-precision force control requires accurate force feedback. The force perception ability of the robotic arm in complex contact operations still faces challenges. Although traditional end force sensors can provide real-time force feedback, in dynamic contact and high-load scenarios, there are problems such as being easily damaged and insufficient accuracy. In addition, the use of a single sensor is difficult to meet the requirements of complex operations for force perception accuracy. It is necessary to fuse data from multiple sensors and use advanced algorithms to achieve accurate identification and compensation of dynamic parameters, and to achieve a large-range measurement and high-precision estimation of the end contact force without using traditional force / torque sensors. And no relatively general measurement technical solutions and hardware systems have been found in relevant domestic and foreign literature. More research has been carried out on specific robotic arms, with low universality and expandability. Summary of the Invention

[0004] To solve the above problems, the present invention discloses a soft measurement system for the contact force at the end of a robotic arm based on the Robot Operating System (ROS). The system consists of four parts: a sensor array, a PLC controller, ROS-related nodes, and a Qt human-computer interaction interface. It realizes the function of automatically measuring the contact force at the tip of the bucket tooth of the excavator robotic arm and displaying it in real time. It reasonably configures the types and quantities of sensors in the contact force measurement system, and places the complex calculation process in the upper computer. It is a relatively general technical solution. PLC + sensors can meet the requirements of most scenarios and functions; borrowing ROS tools simplifies and standardizes the process of soft measurement calculation, standardizes the calculation processes of kinematics and dynamics, and improves the expandability of the measurement system; using Qt realizes the real-time display of measurement data and simulation models, and proposes a 3D visualization solution, comprehensively solving the problem of soft measurement of the contact force at the end of the robotic arm.

[0005] The present invention is a soft measurement system for the contact force at the end of a robotic arm based on the Robot Operating System. The system consists of four parts: a sensor array, a PLC controller, ROS-related nodes, and a Qt human-machine interface. The sensor array transmits signals to the PLC controller in a wired manner and is responsible for collecting temperature, pressure, position, and attitude parameters of the excavator. The PLC controller is responsible for signal conversion and preprocessing of the sensor data and transmits the signals to the ROS-related nodes in the upper computer via the CAN bus. The ROS nodes are responsible for converting, segmenting, processing, and calculating the contact force of the sensor data. The Qt human-machine interface realizes the function of displaying various data transmitted by the ROS nodes and the attitude of the excavator by calling the ROS-related nodes.

[0006] The sensor array consists of three types of sensors: a temperature sensor, a pressure sensor, and an inclination sensor. The temperature and pressure sensors are respectively installed on the bucket, the arm, and the boom. The inclination sensors are respectively installed on the boom, the arm, and the rocker. Among them, the temperature sensor outputs a resistance signal, and the pressure sensor outputs a current signal, which are used to measure the temperature and pressure of the two chambers on both sides of the hydraulic cylinder. There is a piston in the hydraulic cylinder, which divides the entire inner cavity of the hydraulic cylinder into two parts, called the large chamber and the small chamber (or the non-rod side and the rod side). Two temperature and pressure sensors respectively measure the oil temperature and hydraulic pressure of the two chambers on both sides.

[0007] Four sensors are installed on each hydraulic cylinder. The temperature sensor transmits the signal to the resistance-type analog input interface of the PLC controller, and the pressure sensor transmits the signal to the current-type analog input interface of the PLC controller and is then sent to the A / D conversion circuit after processing. The inclination sensor uses an ultra-high-precision single-axis inclination sensor and is installed at the center of gravity of the boom, the arm, and the rocker. Since the position of the bucket is too close to the working point at the end of the excavator's robotic arm and is prone to damage the sensor, it is therefore selected to be installed at the rocker as an alternative, and the angle information of the bucket is obtained through mathematical calculation. The inclination sensor transmits the signal to the CAN transceiver of the PLC controller via the CAN bus. The inclination sensor outputs a digital signal, which is used to measure the current pose of the excavator's robotic arm.

[0008] The PLC controller consists of a current-type analog input processing circuit, a resistance-type analog input processing circuit, an A / D conversion circuit, a CAN transceiver, and a processor. The current-type analog input processing circuit is responsible for receiving the pressure signal, and the resistance-type analog input processing circuit is responsible for receiving the temperature signal. The A / D conversion circuit is used to convert the temperature and pressure signals from analog to digital. One of the two CAN transceivers is responsible for receiving the angle data transmitted by the inclination sensor, and the other is responsible for uploading all the processed and packaged sensor data to the ROS-related nodes in the upper computer via the CAN bus.

[0009] The ROS related nodes are composed of a format conversion node, a data segmentation processing node, a contact force solution node, and an Rviz posture display node in the host computer; the format conversion node is used to receive sensor data transmitted by the PLC controller through the CAN bus, and convert the data format from the standard data frame of the CAN bus to the corresponding data format in ROS; the data segmentation processing node receives the sensor data after format conversion and performs segmentation processing to form separate sensor nodes corresponding to the real sensors; the contact force solution node receives the data of the above-mentioned various sensor nodes for kinematic and dynamic calculations, and solves the joint coordinates and end contact forces in the current motion state; the Rviz posture display node is used to bind the real sensor data to the excavator simulation model, and display the motion state of the real excavator through a 3D simulation model. Rviz is a standard 3D visualization tool in ROS, which can display sensor data, robot status, trajectory planning and environmental models, etc.

[0010] The Qt human-computer interaction interface receives the sensor data from the data segmentation node, the joint coordinates and contact force from the contact force calculation node, and calls the Rviz node to display the excavator motion state. It also has the functions of abnormality detection and data storage. Qt is a cross-platform C++ graphical user interface application development framework that can be jointly developed with ROS.

[0011] The contact force solution node is composed of kinematic calculation, dynamic calculation, load hydraulic rod thrust calculation, no-load hydraulic rod thrust calculation, torque difference calculation and point selection simultaneous equilibrium equations; the kinematic calculation part receives the angle and angular acceleration data transmitted by the boom, dipper arm and rocker angle sensor nodes by subscribing to relevant topics and performs calculations, and then outputs the coordinates of each point at the moment of movement and the angle, angular velocity and angular acceleration data of all joints; the dynamic calculation part receives the angle, angular velocity and angular acceleration data of each joint output by the kinematic calculation and performs calculations, and then outputs the torque of each joint; the no-load hydraulic rod thrust calculation module calculates the torque of each joint according to the dynamics calculation. The calculated torques of each joint are used to calculate the theoretical no-load hydraulic rod thrust; the load hydraulic rod thrust module needs to receive the pressure and temperature sensor node data of the boom, dipper arm, and bucket by subscribing to relevant topics and calculate the actual hydraulic rod thrust in combination with the physical parameters of the hydraulic cylinder; the torque difference calculation module subtracts the no-load hydraulic rod thrust calculated based on the theoretical calculation from the load hydraulic rod thrust calculated based on the actual measured data of the sensor and receives the joint point coordinates to calculate the torque difference of the corresponding joint; the point selection simultaneous equilibrium equation module selects the joint point and the end point to establish the equilibrium equation based on the torque difference and the coordinates of each joint point, and solves the contact force of the end point.

[0012] The load hydraulic rod thrust calculation module consists of temperature data, the pressure in the large chamber of the hydraulic cylinder, the pressure in the small chamber of the hydraulic cylinder, the rod side / non-rod side area, the relationship between the viscous friction coefficient and temperature, and pressure difference calculation. The product of the pressure in the large chamber of the hydraulic cylinder and the non-rod side area gives the original thrust of the large chamber, and the result is output to the pressure difference calculation module. The product of the pressure in the small chamber of the hydraulic cylinder and the rod side area gives the original thrust of the small chamber, and the result is output to the pressure difference calculation module. The relationship between the viscous friction coefficient and temperature is responsible for receiving temperature data and calculating the viscous friction coefficients of the hydraulic oil in the current large and small chambers based on the temperature of the hydraulic cylinder, and the calculation result is output to the pressure difference calculation module. The pressure difference calculation module calculates the actual thrust of the hydraulic rod by receiving the original thrust of the large chamber and the viscous friction coefficient, and the original thrust of the small chamber and the viscous friction coefficient.

[0013] The kinematic calculation module consists of the initial global coordinate system, the initial coordinate transformation matrix, the initial DH parameter table, the initial local coordinate system, the DH parameter table at the movement moment, the inverse coordinate transformation matrix at the movement moment, and the global coordinate system at the movement moment. The initial global coordinate system is the coordinates of each joint point of the excavator manipulator at the initial moment in the earth coordinate system. The initial DH parameter table is a parameter table established based on the position and posture of the excavator manipulator at the initial moment. According to the initial DH parameter table, the transformation matrix of the excavator manipulator in the earth coordinate system and the coordinate system with a certain joint as the coordinate origin at the initial moment can be calculated. The product of the initial global coordinate system and the initial coordinate transformation matrix gives the initial local coordinate system with the corresponding joint point as the coordinate origin. The DH parameter table at the movement moment is established based on the data received from the boom, arm, and rocker inclination sensors, and it represents the basic parameters of the excavator manipulator at that movement moment. According to the DH parameter table at the movement moment, the inverse transformation matrix of the excavator manipulator from the local coordinate system with the corresponding joint point as the coordinate origin to the earth coordinate system at that moment can be calculated. The product of the initial local coordinate system and the inverse coordinate transformation matrix at the movement moment gives the global coordinate system at the movement moment. This coordinate system records the position and posture of the excavator manipulator at the current moment and can directly output the coordinates of each joint point of the manipulator.

[0014] The Qt human-computer interaction interface is composed of starting Qt nodes, graphical human-computer interaction interface, starting data receiving function, starting Rviz function, and starting data storage function; the starting Qt node part is used to run Qt nodes to start the graphical human-computer interaction interface; the graphical human-computer interaction interface is used to start data receiving, Rviz and data storage functions and display the results after starting; the starting data receiving function is to click the data receiving button and Qt will receive sensor data, joint angle data and contact force data and display them in the graphical interface. If the data receiving function is abnormally started, a data receiving error will be reported in the graphical interface; the starting Rviz function is to click the Rviz button and Qt will start the simulation model of the excavator, and bind the real sensor data to the model so that the motion state of the real excavator is displayed on the graphical interface through the 3D model. If the Rviz function is abnormally started, an Rviz function error will be reported in the graphical interface; the starting data storage function is to click the data storage button and Qt will save the sensor data, joint angle data and contact force data received by the data receiving function to the worksheet in the local preset directory in CSV format. If the data storage function is abnormally started, a data storage error will be reported in the graphical interface. This function can only be started after the data receiving function and is invalid before that.

[0015] This measurement system patent includes two parts: hardware circuit and software algorithm. The hardware circuit level is more suitable for the actual working environment of the excavator, and has the characteristics of strong expandability and reusability. Due to the existence of the temperature sensor, the hydraulic oil temperature data can be directly obtained, further improving the modeling accuracy of the force measurement model and the measurement accuracy of the contact force. The upper computer part adds a calculation algorithm for soft measurement of contact force, and uses a series of data to infer (i.e. soft measurement) contact force data, and uses a derivation method to indirectly measure data, avoiding sensor damage caused by drastic changes in hydraulic pressure. At the same time, the use of multi-sensor parameters also improves the indirect measurement accuracy of the measured data. In addition, using ROS as the development environment, the existing tools can be used to simplify the calculation steps of the robot arm, plan the calculation process, and form a modular design to facilitate the expansion and upgrade of the system.

[0016] Beneficial effects of the present invention:

[0017] (1) Modern excavators are usually hydraulic mechanical arm excavators, which have the characteristics of large output torque and high instantaneous impact force. Therefore, it is not reliable to use fragile six-degree-of-freedom torque sensors for contact force measurement. However, the contact force soft measurement based on multiple physical quantities avoids the problem of impact force damaging the sensor. It indirectly measures other physical quantity parameters that are not easy to exceed the range, and then obtains a high-precision estimate based on the relationship between the physical quantities to replace the measured value that is easy to exceed the threshold of the existing sensor. It is extremely safe.

[0018] (2) In this solution, the host computer uses the ROS operating system, which has rich plugins and functional packages, and can relatively easily implement functions such as complex coordinate transformation, robotic arm simulation, and human-machine interaction display. It has strong expandability and portability. The lower computer uses a PLC programmable logic controller, which is more stable and reliable in complex working conditions, and has simple programming, strong flexibility, and good versatility. The combination of ROS + PLC + sensors can meet the software and hardware requirements of most robotic arms, with strong reusability and low development costs. The communication method selects the CAN bus, which has high reliability, real-time performance, and strong anti-interference ability.

[0019] (3) The temperature sensor provides accurate values for the determination of the viscous friction coefficient. The accuracy of the viscous friction force directly affects the measurement accuracy of the contact force. The determination of the viscous friction coefficient can improve the estimation accuracy of the viscous friction force.

[0020] (4) The display of the real-time data of the excavator and the 3D simulation motion model makes the human-machine interaction system more intuitive. The existence of error alarms helps the operator better understand the system status. Brief Description of the Drawings

[0021] Figure 1 is the overall structural schematic diagram of the present invention;

[0022] Figure 2 is the flowchart of the contact force calculation algorithm of the present invention;

[0023] Figure 3 is the flowchart of the load hydraulic rod thrust calculation module of the present invention;

[0024] Figure 4 is the flowchart of the kinematic calculation module of the present invention;

[0025] Figure 5 is the flowchart of the Qt human-machine interaction interface operation of the present invention. Detailed Embodiment

[0026] The following further clarifies the present invention in conjunction with the drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and not to limit the scope of the present invention. It should be noted that the terms "front", "rear", "left", "right", "upper", and "lower" used in the following description refer to the directions in the drawings, and the terms "inner" and "outer" respectively refer to the directions towards or away from the geometric center of a specific component.

[0027] Figure 1This is the overall structural schematic diagram of the present invention. This embodiment is a soft measurement system for the contact force at the end of a robotic arm based on the Robot Operating System (ROS). The system consists of four parts: a sensor array, a PLC controller, ROS-related nodes, and a Qt human-machine interface. The sensor array transmits signals to the PLC controller via a wired connection. The PLC controller transmits signals to the ROS-related nodes in the upper computer via a wired connection. After the nodes process and calculate the data, the contact force results and the pose of the excavator are sent to the Qt human-machine interface. The sensor array consists of three types of sensors: a temperature sensor, a pressure sensor, and an inclination sensor. The pressure and temperature sensors are installed in three parts: the boom, the arm, and the bucket. The temperature sensor outputs a resistance signal, and the pressure sensor outputs a current signal. Two pressure sensors and two temperature sensors are installed on each hydraulic cylinder to measure the hydraulic pressure on both sides of the hydraulic cylinder cavity and the hydraulic oil temperature. The pressure and temperature sensors transmit the signals to the corresponding interfaces of the PLC controller, and the signals are converted into digital signals through the processing circuit and A / D conversion circuit inside the PLC controller. The inclination sensor is installed at the center of gravity of the boom, the arm, and the rocker, and transmits the signal to the CAN transceiver inside the PLC controller. The inclination sensor outputs a digital signal and is used to measure the current pose of the excavator robotic arm.

[0028] The PLC controller consists of an analog input processing circuit, an A / D conversion circuit, a CAN transceiver, and a processor. It is used to uniformly convert the signals of each sensor into digital signals, process and package the data, and send it to the upper computer through the CAN transceiver.

[0029] The ROS-related nodes consist of a format conversion node, a data segmentation node, a contact force calculation node, and an Rviz pose display node in the upper computer. The format conversion node is used to receive the data transmitted by the PLC controller through the CAN bus and convert the data format from the standard data frame of the CAN bus to the corresponding data format in ROS. The data segmentation node receives the sensor data after format conversion and performs segmentation processing to form individual sensor nodes corresponding to the real sensors. The contact force calculation node receives the data of the above various sensor nodes for kinematic and dynamic calculations, and calculates the joint coordinates and the end contact force in the current motion state. The Rviz pose display node is used to bind the real sensor data to the excavator simulation model and display the motion state of the real excavator through the 3D simulation model.

[0030] The Qt human-machine interface receives the sensor data transmitted by the data segmentation node, the joint coordinates and the contact force transmitted by the contact force calculation node, and calls the Rviz node to display the motion state of the excavator. At the same time, it also has the functions of anomaly detection and data storage.

[0031] Figure 2This is the flow chart of the contact force calculation algorithm of the present invention. The contact force calculation node consists of kinematic calculation, dynamic calculation, load hydraulic rod thrust calculation, no-load hydraulic rod thrust calculation, torque difference calculation, and point selection and simultaneous equilibrium equation; the kinematic calculation part receives the angle and angular acceleration data transmitted by the boom, stick, and rocker inclination sensor nodes by subscribing to relevant topics and performs operations, and then outputs the coordinates of each point at the moment of movement and the angles, angular velocities, and angular acceleration data of all joints; the dynamic calculation part receives the angles, angular velocities, and angular acceleration data of each joint output by the kinematic calculation and performs operations, and then outputs the torque of each joint; the no-load hydraulic rod thrust calculation module calculates the no-load hydraulic rod thrust in the theoretical case according to the torque of each joint obtained by the dynamic calculation; the load hydraulic rod thrust module needs to receive the data of the pressure and temperature sensor nodes of the boom, stick, and bucket by subscribing to relevant topics and calculate the actual hydraulic rod thrust in combination with the physical parameters of the hydraulic cylinder; the torque difference calculation module subtracts the no-load hydraulic rod thrust calculated according to the theoretical calculation from the load hydraulic rod thrust calculated according to the actual measured data of the sensor and receives the joint point coordinates to calculate the torque difference of the corresponding joint; the point selection and simultaneous equilibrium equation module selects joint points and end points to establish an equilibrium equation according to this torque difference and the coordinates of each joint point, and solves the contact force of the end point;

[0032] Figure 3 This is the flow chart of the load hydraulic rod thrust calculation module of the present invention. The load hydraulic rod thrust calculation module consists of temperature data, large chamber pressure of the hydraulic cylinder, small chamber pressure of the hydraulic cylinder, rod side / non-rod side area, relationship between viscous friction coefficient and temperature, and pressure difference calculation; multiplying the large chamber pressure of the hydraulic cylinder by the non-rod side area gives the original thrust of the large chamber, and the result is output to the pressure difference calculation module; multiplying the small chamber pressure of the hydraulic cylinder by the rod side area gives the original thrust of the small chamber, and the result is output to the pressure difference calculation module; the relationship between viscous friction coefficient and temperature is responsible for receiving the temperature data and calculating the viscous friction coefficient of the hydraulic oil in the current large and small chambers according to the temperature of the hydraulic cylinder, and outputting the calculation result to the pressure difference calculation module; the pressure difference calculation module calculates the thrust of the actual hydraulic rod by receiving the original thrust of the large chamber and the viscous friction coefficient, and the original thrust of the small chamber and the viscous friction coefficient;

[0033] Figure 4This is the flowchart of the kinematic calculation module of the present invention. The kinematic calculation module consists of an initial global coordinate system, an initial coordinate transformation matrix, an initial DH parameter table, an initial local coordinate system, a DH parameter table at the moment of motion, an inverse coordinate transformation matrix at the moment of motion, a global coordinate system at the moment of motion, and a coordinate output module; the initial global coordinate system is the coordinates of each joint point of the excavator manipulator at the initial moment in the earth coordinate system; the initial DH parameter table is a parameter table established according to the position and attitude of the excavator manipulator at the initial moment, and according to the initial DH parameter table, the transformation matrix of the excavator manipulator in the earth coordinate system and the coordinate system with a certain joint as the coordinate origin at the initial moment can be calculated; the initial global coordinate system is multiplied by the initial coordinate transformation matrix to obtain the initial local coordinate system with the corresponding joint point as the coordinate origin; the DH parameter table at the moment of motion is established according to the data received from the boom, stick, and rocker inclination sensors, and it represents the basic parameters of the excavator manipulator at this moment of motion; according to the DH parameter table at the moment of motion, the inverse transformation matrix of the excavator manipulator from the local coordinate system with the corresponding joint point as the coordinate origin to the earth coordinate system can be calculated; the initial local coordinate system is multiplied by the inverse coordinate transformation matrix at the moment of motion to obtain the global coordinate system at the moment of motion; this coordinate system records the position and attitude of the excavator manipulator at the current moment, and the coordinates of each joint point of the manipulator can be directly output;

[0034] Figure 5 This is the flowchart of the operation of the Qt human-machine interaction interface of the present invention. The Qt human-machine interaction interface consists of a Qt node startup, a graphical human-machine interaction interface, a data reception function startup, an Rviz function startup, and a data storage function startup; the Qt node startup part is used to run the Qt node to start the graphical human-machine interaction interface; the graphical human-machine interaction interface is used for data reception, the opening of the Rviz and data storage functions, and the display of the results after opening; the role of the data reception function startup is that after clicking the data reception button, Qt will receive sensor data, joint angle data, and contact force data and display them in the graphical interface. If the data reception function starts abnormally, a data reception error will be reported in the graphical interface; the role of the Rviz function startup is that after clicking the Rviz button, Qt will start the simulation model of the excavator and bind the real sensor data to the model so that the motion state of the real excavator is displayed in the graphical interface through the 3D model. If the Rviz function starts abnormally, an Rviz function error will be reported in the graphical interface; the role of the data storage function startup is that after clicking the data storage button, Qt will save the sensor data, joint angle data, and contact force data received by the data reception function to a worksheet in a local preset directory in the CSV format. If the data storage function starts abnormally, a data storage error will be reported in the graphical interface. This function can only be started after the data reception function and is invalid before that.

[0035] The technical means disclosed by the solution of the present invention are not limited to the technical means disclosed in the above embodiments, but also include technical solutions composed of any combination of the above technical features.

Claims

1. A soft measurement system for contact force at the end of a manipulator based on a robot operating system, the system consists of a sensor array, a PLC controller, ROS-related nodes and a Qt human-computer interaction interface; the sensor array transmits signals to the PLC controller by wire; the PLC controller transmits signals to the ROS-related nodes in the host computer by wire; the ROS-related nodes process and solve the data and send the contact force results and the excavator posture to the Qt human-computer interaction interface; the system is characterized in that: The sensor array is composed of three types of sensors, namely, temperature sensor, pressure sensor and inclination sensor; the pressure sensor and temperature sensor are respectively installed on the boom, the dipper rod and the bucket; the temperature sensor outputs a resistance signal, the pressure sensor outputs a current signal, and each hydraulic cylinder is installed with two pressure sensors and temperature sensors, which are respectively used to measure the hydraulic pressure of the cavity on both sides of the hydraulic cylinder and the temperature of the hydraulic oil; the pressure and temperature sensors transmit the signals to the corresponding interface of the PLC controller, and are converted into digital signals through the processing circuit and A / D conversion circuit inside the PLC controller; the inclination sensor is installed at the center of gravity of the boom, the dipper rod and the rocker, and transmits the signal to the CAN transceiver inside the PLC controller, and the inclination sensor outputs a digital signal for measuring the current posture of the excavator mechanical arm; the PLC controller is composed of an analog input processing circuit, an A / D conversion circuit, a CAN transceiver and a processor; it is used to uniformly convert the sensor signals of each channel into digital signals and process and package the data and send them to the host computer through the CAN transceiver; the ROS related nodes are composed of a format conversion node, a data segmentation node, a contact force solution node and an Rviz posture display node in the host computer.

2. According to claim 1, a robot arm end contact force soft measurement system based on a robot operating system is characterized in that: The format conversion node is used to receive data transmitted by the PLC controller through the CAN bus, and convert the data format from the standard data frame of the CAN bus to the corresponding data format in ROS; the data segmentation node receives the sensor data after format conversion and performs segmentation processing to form separate sensor nodes corresponding to the real sensors; The contact force solution node receives data from the above-mentioned sensor nodes for kinematic and dynamic calculations, and solves the joint coordinates and end contact forces in the current motion state; the Rviz posture display node is used to bind the real sensor data to the excavator simulation model, and display the motion state of the real excavator through the 3D simulation model; the Qt human-computer interaction interface receives the sensor data transmitted by the data segmentation node, the joint coordinates and contact force transmitted by the contact force solution node, and calls the Rviz node to display the motion state of the excavator. It also has the functions of abnormality detection and data storage.

3. The soft measurement system for contact force of a manipulator end based on a robot operating system according to claim 1, characterized in that: The contact force solution node is composed of kinematic calculation, dynamic calculation, load hydraulic rod thrust calculation, no-load hydraulic rod thrust calculation, torque difference calculation and point selection simultaneous equilibrium equations; the kinematic calculation part receives the angle and angular acceleration data transmitted by the boom, dipper arm and rocker tilt sensor nodes by subscribing to relevant topics and performs calculations, and then outputs the coordinates of each point at the moment of movement and the angle, angular velocity and angular acceleration data of all joints; the dynamic calculation part receives the angle, angular velocity and angular acceleration data of each joint output by the kinematic calculation and performs calculations, and then outputs the torque of each joint; the no-load hydraulic rod thrust calculation module calculates the no-load hydraulic rod thrust under theoretical conditions according to the torque of each joint obtained by the dynamic calculation; The load hydraulic rod thrust module needs to receive the pressure and temperature sensor node data of the boom, dipper arm, and bucket by subscribing to relevant topics and calculate the actual hydraulic rod thrust in combination with the physical parameters of the hydraulic cylinder; the torque difference calculation module subtracts the no-load hydraulic rod thrust calculated based on the theoretical calculation and the load hydraulic rod thrust calculated based on the actual measured data of the sensor and receives the joint point coordinates to calculate the torque difference of the corresponding joint; the point selection simultaneous equilibrium equation module selects the joint point and the end point according to the torque difference and the coordinates of each joint point to establish the equilibrium equation and solve the contact force of the end point.

4. The soft measurement system for contact force of a manipulator end based on a robot operating system according to claim 1, characterized in that: The load hydraulic rod thrust calculation module is composed of temperature data, hydraulic cylinder large chamber pressure, hydraulic cylinder small chamber pressure, rod side / non-rod side area, viscous friction coefficient and temperature relationship, and pressure difference calculation; the hydraulic cylinder large chamber pressure is multiplied by the non-rod side area to obtain the large chamber original thrust, and the result is output to the pressure difference calculation module; the hydraulic cylinder small chamber pressure is multiplied by the rod side area to obtain the small chamber original thrust, and the result is output to the pressure difference calculation module; the viscous friction coefficient and temperature relationship is responsible for receiving temperature data and calculating the viscous friction coefficient of the current large and small chamber hydraulic oil according to the hydraulic cylinder temperature, and outputting the calculation result to the pressure difference calculation module; the pressure difference calculation module calculates the actual hydraulic rod thrust by receiving the large chamber original thrust and viscous friction coefficient, the small chamber original thrust and viscous friction coefficient.

5. The soft measurement system for contact force of a manipulator end based on a robot operating system according to claim 1, characterized in that: The kinematic calculation module is composed of an initial global coordinate system, an initial coordinate transformation matrix, an initial DH parameter table, an initial local coordinate system, a DH parameter table at the motion moment, an inverse matrix of coordinate transformation at the motion moment, a global coordinate system at the motion moment, and a coordinate output module; the initial global coordinate system is the coordinates of each joint point of the excavator mechanical arm in the earth coordinate system at the initial moment; the initial DH parameter table is a parameter table established according to the position and posture of the excavator mechanical arm at the initial moment, and the transformation matrix of the excavator mechanical arm in the earth coordinate system and the coordinate system with a certain joint as the coordinate origin at the initial moment is calculated according to the initial DH parameter table; the initial global coordinate system and the initial The initial local coordinate system with the corresponding joint point as the coordinate origin is obtained by multiplying the initial coordinate transformation matrix; the DH parameter table at the movement moment is established according to the received boom, arm and rocker tilt sensor data, which represents the basic parameters of the excavator mechanical arm at the movement moment; the inverse matrix of the transformation of the excavator mechanical arm from the local coordinate system with the corresponding joint point as the coordinate origin to the earth coordinate system is calculated according to the DH parameter table at the movement moment; the global coordinate system at the movement moment is obtained by multiplying the initial local coordinate system with the inverse matrix of the coordinate transformation at the movement moment; this coordinate system records the position and posture of the excavator mechanical arm at the current moment, and can directly output the coordinates of each joint point of the mechanical arm.

6. The soft measurement system for contact force of a manipulator end based on a robot operating system according to claim 1, characterized in that: The Qt human-computer interaction interface consists of starting the Qt node, the graphical human-computer interaction interface, starting the data receiving function, starting the Rviz function, and starting the data storage function; the starting Qt node part is used to run the Qt node to start the graphical human-computer interaction interface; the graphical human-computer interaction interface is used to start the data receiving, Rviz and data storage functions and display the results after starting; The function of starting the data receiving function is that after clicking the data receiving button, Qt will receive the sensor data, joint angle data and contact force data and display them in the graphical interface. If the data receiving function is started abnormally, a data receiving error will be reported in the graphical interface; The function of starting the Rviz function is that after clicking the Rviz button, Qt will start the simulation model of the excavator and bind the real sensor data to the model so that the motion state of the real excavator can be displayed on the graphical interface through the 3D model. If the Rviz function is started abnormally, an Rviz startup error will be reported in the graphical interface. The function of starting the data storage function is that after clicking the data storage button, Qt will save the sensor data, joint angle data and contact force data received by the data receiving function to the worksheet in the local preset directory in the format of CSV; if the data storage function is started abnormally, a data storage error will be reported in the graphical interface. This function can only be started after the data receiving function, and it is invalid before that.

Images