When ROV (Remotely Operated Vehicle) moves, its precise mathematical model is difficult to establish, and it often fails to achieve the consistency between motion and control. The design of ROV remote cooperative controller based on multi-algorithm includes two parts: ROV remote communication module and multi-algorithm cooperative controller.
The design of ROV remote communication adopts serial port conversion Ethernet module communication mode, which transmits the collected sensor information to the upper computer in JSON data format in real time. The multi-algorithm cooperative controller includes ROV yaw and depth feed-forward PID controller and ROV pitch sliding mode controller, which can realize the state control of the robot itself. Finally, the corresponding control effect and experimental parameters are obtained through a number of experiments, which verify the effectiveness and practicability of the remote cooperative controller. With the development of the times, the application fields of underwater vehicles in the world are expanding, including ocean research (scientific investigation), ocean development and underwater engineering, and have developed rapidly in many countries. The military powers represented by the United States attach great importance to the application of underwater vehicles in future wars. In the process of ocean development, underwater vehicles will play an important role in the detection and modeling of complex marine environment, as well as underwater detection and recognition, positioning and transmission of ocean targets. How to effectively control the motion and attitude of underwater vehicles is a problem that many researchers attach great importance to. Because autonomous underwater vehicle (AUV) motion control has the characteristics of time-varying, non-linearity, large time-delay and multi-degree of freedom, and because of the complexity and variability of the underwater environment, it is difficult to obtain an accurate system model, which poses a great challenge to the precise positioning and control of AUV. With the expanding application of autonomous underwater vehicle, one of the key technologies to improve its autonomy is to improve the performance of its controller system [1]. In recent years, underwater vehicles (AUVs) have developed rapidly. However, due to their various operation forms, the shape of AUVs is also different, and they usually do not have streamlined structure. However, the existing several typical mathematical models are built for a specific object, so there is no general model for underwater vehicles. In addition, due to the large underwater disturbance (ocean current), the difficulty of fixed-point control (dynamic positioning) of underwater vehicles is further increased. Among all kinds of control algorithms for robots, sliding mode control is one of the effective control schemes for underwater vehicles because of its robustness to parameter changes and unmodeled parts of the system and strong anti-interference. However, the existing models are designed for different objects, and how to mitigate the coupling effect caused by the applied control effect has not been studied. Therefore, the hovering control algorithm and direction research of underwater vehicle is a very valuable research topic. In view of this, the corresponding controller is designed in this paper. The depth and course of ROV are controlled by PID algorithm, and the pitch attitude of ROV is controlled by sliding mode algorithm. Among them, the algorithm design and parameter tuning have been simulated in MATLAB, and have been further verified in the actual test. Fig. 1 is an underwater vehicle designed in this paper.
For the pitch attitude control of ROV with small angle, the PID linear control method is usually used to solve the problem. However, when the pitch attitude angle of ROV is greater than 15 degrees, the PID algorithm will fail and the pitch attitude of ROV will oscillate. The pitching attitude of ROV is similar to that of simple pendulum model and belongs to non-linear control problem. Therefore, this paper designs a sliding mode control algorithm to overcome this non-linear control problem. ROV remote communication is the hub for underwater vehicle to communicate with the ground while performing underwater tasks. This paper uses USR-WIFI232 module and power line carrier module to feedback the data collected by the underwater vehicle to the water surface in real time. The data format of communication is the basis of the whole algorithm. Therefore, choosing an effective data exchange format is the key to the design. In this paper, JSON data format is used to solve the problems of sensor data parsing and generation on different platforms. JSON (JavaScript Object Notation) is a lightweight data exchange format.
It is based on a subset of ECMAScript and adopts a completely language-independent text format, but also uses habits similar to the C language family (including C, C , C#, Java, JavaScript, Perl, Python, etc.). It is easy to read and write, and easy to machine parse and generate (generally used to improve network transmission rate). These features make JSON an ideal data exchange language [3]. Therefore, ROV system uses JSON communication format written in C language (see figure 2) to send sensor data to PC in real time. On various software platforms (MATLAB, LABVIEW, etc.), SDK parses the data packages in this format, which facilitates user development.
The USR-WIFI232-D2 module is an integrated 802.11b/g/n WiFi and Ethernet module. It provides a solution for connecting users’physical devices to WiFi wireless network or Ethernet network and providing USART data transmission interface. Through this module, traditional low-end serial devices or MCU-controlled devices can easily access WiFi wireless network or ethernet, thus realizing the control and management of the Internet of Things network [4]. Among them, motors object is represented in JSON as the content in “{}”, and data structure is the content in the second layer “{}”. In object-oriented language, set_motor1_speed is the attribute of motor 1 and 0 is the corresponding attribute value, so it is easy to understand. The analytical method is motors. set_motor1_speed to get the attribute value. The type of attribute value can be number, string, array, object, etc. So JSON not only improves readability, but also reduces complexity and is easy to program.
It only needs to satisfy the JSON format. In formula (1), P (t) is the output signal of the regulator; e (t) is the deviation signal of the regulator, which equals the difference between the measured value and the given value; KP is the proportional coefficient of the regulator; KI is the integral time of the regulator; and KD is the differential time of the regulator. Because computer control is a kind of sampling control, it can only calculate the control quantity according to the deviation value of sampling time. Therefore, in the computer control system, the first step is to discretize the equation (1) and replace the differential equation [5] of the continuous system with the difference equation in digital form. It can be seen from formula (4) that to calculate P (k), not only the deviation signals E (k) and E (k-1) of this time and the last time, but also the previous deviation signals E (j) are added in the integral term, i.e., the PID algorithm of the discrete system is obtained from NJ = 0E (j). However, in the PID control algorithm applied to the underwater vehicle, there will always be some steady-state errors and delays in the course and depth control of the robot itself. Therefore, some improvements have been made in this paper. After the output of PID, adding a feed-forward control quantity (depth and heading angle) can speed up the control response and eliminate the steady-state error in time. In addition, there are some differences between depth control and course control. In the process of attitude sensor, yaw of course is – 180 ~180. In this process, a shortest path algorithm can be used to make the underwater vehicle reach the set course angle with the shortest path and time. In the pool experiment, the realization of the course angle control of the underwater robot based on PID is discussed. When the absolute value of the difference between the actual angle and the set value of the underwater vehicle is greater than 90 degrees, a feed-forward heading angle is given to the robot by feedforward control to make the robot respond to the set value immediately. When the absolute value of the difference is less than or equal to 90 degrees, the robot will automatically adopt the PID controller to adjust the output to achieve the corresponding set value (allowable error range). Internal). Following is how the program realizes the process of “feedforward feedback” control of underwater vehicle through PID control algorithm. Firstly, it receives the data of the actual angle yaw of the robot itself by interrupting the serial port, and sends the data to the upper computer of MATLAB. Secondly, it judges whether the absolute difference between the set value and yaw is greater than 90 degrees at the moment, so as to decide. After the output of the PID, the feed-forward heading angle is added. Finally, the output speed is sent to the BLDCM through I2C to change the angle of the robot itself to approach the set target angle [7]. The program block diagram is shown in Figure 3. Among them, I denotes the moment of inertia, theta, theta. denotes pitch, angular velocity and angular acceleration of the underwater vehicle, respectively. M denotes the mass of the underwater vehicle, L2 denotes the vertical distance from the center of gravity to the center of buoyancy, thermostatic element F1 and F2 denote the motor moment, l denotes the vertical distance from the motor moment to the center of buoyancy. The switched hyperplane is designed according to the desired dynamic characteristics of the underwater vehicle system, and the system states are switched from the hyperplane to the switched hyperplane by sliding mode controller. Once the system reaches the switching hyperplane, the control function will ensure that the system reaches the system origin along the switching hyperplane. Theta and theta-gamma represent the measured pitch and the set pitch respectively, and P represent the parameters that need to be adjusted. F, Ki, A and B represent the parameters that need to be adjusted. These parameters can be adjusted to meet the arrival conditions, so that the approaching motion (non-sliding mode) reaches the switch surface in a finite time. In the process of reaching, it has the characteristics of fast speed and small chattering, can overcome the uncertainty of the system, and has strong robustness to the complexity of hydrodynamic coefficients and the change of inrush environment, so that the pitch angle control of the underwater vehicle achieves good results. Compared with the PID control, the pitch angle control process belongs to the non-linear system control, and the PID control is an algorithmic controller for the linear system design, which is more sensitive to interference [811]. Therefore, when the controlled object is controlled by sliding mode control, it has strong robustness to the model error, parameter variation and external disturbance of the controlled object, and can be widely used in complex practical situations. The block diagram of the algorithm is shown in Figure 4. According to the yaw and roll model of underwater vehicle, the feedforward PID parameters can be obtained by analyzing the experimental data and simulation data, as shown in Table 1. The simulation and experimental test of pitch angle, course control of underwater vehicle are shown in Fig. 5 and Fig.
6. The fine line represents the course and depth control of the robot. The data of attitude sensor (IMU) and pressure sensor are obtained, and the thick line represents the graphics simulated by MATLAB. Since the sensor itself has an error of about 2 degrees or 2 mv, the actual waveform will fluctuate within the allowable error range. The stability and anti-interference ability of ROV is the key index of robot in practice, and the biggest underwater interference is current, which further increases the difficulty of ROV fixed-point control. In this paper, the robot is put into a river with a certain velocity of water in the simulated ocean current experiment. The pitch angle of the robot is set at – 15 degrees, and the robot itself controls the pitch angle of the robot by sliding mode controller to – 15 degrees. By observing the elevation data curve, it is found that ROV has good stability and anti-jamming ability, which verifies the practicability and effectiveness of the algorithm. In the simulated ocean current experiment, the yaw angle data curve of Fig. 6 is observed. The ROV yaw angle has fast dynamic response, high system stability and strong anti-jamming ability, which achieves the consistency of robot motion and manipulation. The feedforward PID control is used in the yaw control of the robot, and the parameter adjustment is simple, so the sliding mode control algorithm is not used. The yaw angle fluctuation in the figure belongs to the problem of sensor accuracy, which can be neglected in the normal error range. ROV takes water surface as reference voltage to represent 0 mv, abscissa represents time, and ordinate represents voltage value after calibration of pressure sensor. With the increase of ROV diving depth, the sensor voltage increases. The curve in Figure 7 describes the curve of ROV which is stable at the set value of 3 MV under the control of feed-forward PID. The curve fluctuation is within the normal error range and can be neglected. In this paper, the motion control of the underwater vehicle is analyzed. The state information of the underwater vehicle itself is fed back to the host computer through the serial port switching Ethernet module, and the motion control of the underwater vehicle is carried out by using the MATLAB platform.