CN105527960A - Mobile robot formation control method based on leader-follow - Google Patents

Abstract

A mobile robot formation control method based on pilot following, which is composed of a global positioning system, a wireless communication system, an algorithm processing system, and a speed control system. The global positioning system obtains the pose information of each robot, and sends it to the algorithm for processing through the wireless communication system System, through the information interaction with the speed control system, the formation motion control is finally realized. In the control algorithm, first establish the motion model of the leader following the formation, give the motion control rate of the following robot, then establish the trajectory prediction model of the following robot, use the nonlinear least square method to predict the model, use the improved particle swarm optimization algorithm to optimize the prediction model parameters, and define the communication Data anomaly range, enabling prediction points instead of anomalous points to ensure formation movement. The invention introduces a predictive model to avoid formation deviation caused by temporary communication abnormalities, ensure the reliability of following the robot movement, and greatly improve the stability of the formation.

Description

A mobile robot formation control method based on leader following

technical field

The invention relates to the field of motion control of mobile robots, in particular to a method for controlling the motion of mobile robot formations based on pilot follow.

Background technique

With the development of robot technology, mobile robots have been greatly expanded and improved in terms of application scope and functions, and are widely used in national defense, industry, service industry and other industries. The formation of multiple mobile robots can effectively make up for the lack of ability of a single mobile robot, so that a single robot has a significant advantage in the group of formation behaviors that it did not have before. Through the coordination and cooperation of robots between groups, it can complete the work that a single robot cannot or is difficult to complete. Therefore, it is very necessary to study the formation control system and control method of mobile robots.

After searching the existing literature, the Chinese patent application number is: CN201010618568.1, and the name is: multi-robot formation method based on Ad-Hoc network and leader-follower algorithm. This invention proposes a multi-robot formation method to establish a leader The kinematics model determines the follower's trajectory according to the artificial potential field method, and introduces the Ad-Hoc network between the leader and the follower, and establishes information feedback to ensure that the follower's tracking process of the leader is not lost. However, the invention introduces the Ad-Hoc network into the pilot follower and does not consider the abnormal network communication, which is prone to follower movement deviation phenomenon caused by abnormal network communication. Chinese patent application number: 201310281499.3, titled: A behavior-based leader-follower multi-agent formation control method, the invention discloses a behavior-based leader-follower multi-agent formation control method , this method makes up for the defect that the leader leads to formation failure after an obstacle occurs to a certain extent by adding alternate leaders. However, this invention does not consider the problem that the follower fails to obtain the leader's information in the lead-follow formation, which may easily lead to the loss of the follower. The name of the document is: Design of a new controller for multi-agent pilot-follow formation. In this algorithm, a new controller is designed by introducing neighbor-based local control rate and neighbor-based state estimation rules. In this controller Arbitrary formations can be easily realized by simply setting the relative coordinates between the leader and the follower, but this method forms a fixed or switchable connection topology for the formation composed of the leader and follower, which lacks certain formation flexibility And it does not take into account the problem of leader-following formation under the condition of limited communication topology.

Through the search, it is found that there are many studies on the motion control of mobile robots leading and following formations in the prior art. Although all of them can realize the formation control of robots, they have not considered the problem of follower movement in the case of limited communication in leading and following formations. , poor formation stability, lack of effective formation protection. Therefore, the present invention proposes a mobile robot formation control method based on leader following.

Contents of the invention

The purpose of the present invention is to provide a mobile robot formation control method based on leader following that is obviously optimized for communication anomalies, follower movement conditions and overall formation formation maintenance in leader following formation movement.

In order to achieve the above object, the following technical solutions are adopted: the control method of the present invention includes a global positioning system, a wireless communication system, an algorithm processing system, a speed control system, and mobile robots of different colors; The mobile robot performs color recognition and collects the pose information of the mobile robot; the wireless communication system sends the collected pose information to each robot; the algorithm processing system makes motion control and coordination formation algorithms based on the received pose information of each mobile robot The speed control system collects and processes the speed information of each mobile robot, continuously makes feedback on the actual control output, corrects the speed deviation in real time, and finally completes the control of the formation movement of each mobile robot.

The concrete steps of described control method are as follows:

Step 1, on the working platform, use the global positioning system to obtain the position coordinate information of each robot in the formation relative to the global coordinate system of the platform;

Step 2. According to the position coordinate information of each robot, construct the relative position angle relationship between the leader robot and the follower robot, define the distance angle relationship between the leader robot and the follower robot expected by the formation, obtain the error dynamics equation, and then derive the follower robot motor control rate;

Step 3. During the formation movement, the leader robot moves autonomously to the target point and continuously sends information to the follower robot through the wireless communication system. The follower robot receives the information of the leader robot and calculates its own position, and records the position information of the follower robot at each moment as Training data, establish a trajectory prediction model based on the nonlinear least squares method to follow the robot;

Step 4, use the particle swarm algorithm to solve the problem of parameter selection of the trajectory prediction model, use a set of parameters of the trajectory prediction model as a set of solutions of the trajectory prediction model, initialize the particle swarm, calculate the fitness function of each particle, according to the particle fitness The function value updates the individual extremum and the global extremum of the particle;

Step 5, using an iterative formula to update the velocity and position of each particle;

Step 6. Continuously update the trajectory prediction model with the data at the next moment, and predict the trajectory information backwards, and define an abnormal range of communication data. If the received data is within the normal range, calculate the following based on the received pilot robot data. Robot position information and move towards it; if the data is abnormal, enable the prediction point instead of the abnormal data point as the following robot trajectory point at the next moment;

Step 7, use the global positioning system to continuously update the position and attitude information of each robot in the formation until the leader follows the formation to reach the target point.

Further, in step 6, the abnormal range of the communication data is to use the global positioning system and the wireless communication system to collect each frame of image, and calculate the time interval for sending the robot position information and the average moving speed of the robot to determine whether the communication data is abnormal data.

Further, in step 6, the data update strategy is that the trajectory point value of the following robot at the next moment is determined according to the relationship between the trajectory point value of the following robot at the current moment and the maximum threshold value of the data exceeding the normal range, and by separately judging Update decisions are made based on whether the distance between the current and next robot trajectory point values along the X and Y axes of the platform meets the abnormal range of communication data.

Further, the global positioning system is composed of a host computer and a high-definition CCD camera. The high-definition CCD camera is placed on the top of the work platform, and each mobile robot in the platform is marked with a different color, and the color recognition algorithm is used to obtain the position of the robot on the platform. The position and attitude information in the mobile robot is then sent back to the host computer for processing, and then sent to each mobile robot after processing by the host computer, so as to realize the global positioning of the mobile robot.

Further, the wireless communication system includes a host computer, a mobile robot and a wireless router, wherein a wireless transceiver module is installed on the mobile robot, the host computer and the wireless router are connected through a network cable, and the host computer, the mobile robot and the wireless router form a network based on a local area network. The communication system realizes wireless communication between the host computer and robots, and between robots by establishing communication protocols between the host computer and each robot and between each robot.

Further, the algorithm processing system is composed of a DSP digital signal processor, which is installed inside the robot, and the aforementioned pilot-following-based mobile robot formation motion control algorithm is implanted in the DSP digital signal processor. The data of the communication system is used to process the robot's autonomous movement and formation coordination control;

Further, the speed control system includes a speed measuring code disc module, an Arduino module and a PID speed control module;

Wherein, the speed measuring code disc module is composed of a grating disc and a photoelectric detection device, installed beside the left and right wheel motors of the robot, and reflects the current revolutions of the left and right wheel motors by calculating the number of output pulses of the photoelectric encoder per second, and combining The circumference of a wheel rotation is converted into the actual speed of the left and right wheels of the mobile robot motor;

The Arduino module is an ArduinoMega2560 single-chip microcomputer, which is installed inside the robot, and is used to expand the digital and analog input and output ports of the robot, and then collect and receive the output pulse number of the speed measuring code disc. After analyzing and processing the data, it sends it to the algorithm through the I2C bus. processing system;

The PID speed control module is a PID controller based on the BP neural network. On the basis of the classic PID controller, the measured speed returned by the Arduino module is compared with the theoretical speed in the algorithm processing system. The network optimizes the PID controller parameters, so as to realize the adaptive adjustment of the controller parameters of the robot in different motion states, and correct the speed deviation in real time.

Compared with the prior art, the present invention has the following advantages: the trajectory prediction model of the following robot is established, and the least squares prediction model is optimized by using the improved particle swarm optimization algorithm, which makes up for the formation deviation phenomenon caused by short-term communication abnormalities, and ensures It improves the reliability of the follower movement and greatly improves the stability of the formation.

Description of drawings

Fig. 1 is a system structure block diagram of the method of the present invention.

Fig. 2 is a flow chart of the control method of the method of the present invention.

Reference numerals: 1 is a global positioning system, 2 is a wireless communication system, 3 is an algorithm processing system, 4 is a speed control system, 41 is a speed measuring code disc module, 42 is an Arduino module, and 43 is a PID speed control module.

detailed description

The present invention will be further described below in conjunction with accompanying drawing:

As shown in Figure 1, the control method of the present invention includes a global positioning system 1, a wireless communication system 2, an algorithm processing system 3, a speed control system 4, and mobile robots of different colors; The robot performs color recognition and collects the pose information of the mobile robot; the wireless communication system sends the collected pose information to each robot in the form of UPD broadcast; the algorithm processing system makes motion control according to the received pose information of each mobile robot And coordinate formation algorithm processing, and constantly interact with the speed control system, the speed control system collects and processes the speed information of each mobile robot, continuously makes feedback on the actual control output, corrects the speed deviation in real time, and finally completes the formation movement of each mobile robot control.

Wherein, the global positioning system is composed of a host computer and a high-definition CCD camera. The high-definition CCD camera is placed on the top of the work platform, and each mobile robot in the platform is marked with a different color. The position and attitude information of the mobile robot is then transmitted back to the host computer for processing, and then sent to each mobile robot after processing by the host computer, so as to realize the global positioning of the mobile robot.

The wireless communication system includes a host computer, a mobile robot and a wireless router, wherein the mobile robot is equipped with a wireless transceiver module, and the host computer, the mobile robot and the wireless router form a communication system based on a local area network. By establishing a communication system between the host computer and each robot And the communication protocol between the robots to realize the wireless communication between the upper computer and the robot, and between the robot and the robot.

The algorithm processing system is composed of a DSP digital signal processor. In the DSP digital signal processor, a mobile robot formation motion control algorithm based on pilot following is implanted, receiving data from a wireless communication system, and controlling the autonomous movement and formation of the robot. Coordinated control to handle;

The speed control system includes a speed measuring code disc module 41, an Arduino module 42 and a PID speed control module 43;

Wherein, the speed measuring code disc module is composed of a grating disc and a photoelectric detection device, installed beside the left and right wheel motors of the robot, and reflects the current revolutions of the left and right wheel motors by calculating the number of output pulses of the photoelectric encoder per second, and combining The circumference of a wheel rotation is converted into the actual speed of the left and right wheels of the mobile robot motor;

The Arduino module is an ArduinoMega2560 single-chip microcomputer, which is installed inside the robot, and is used to expand the digital and analog input and output ports of the robot, and then collect and receive the output pulse number of the speed measuring code disc. After analyzing and processing the data, it sends it to the algorithm through the I2C bus. processing system;

The PID speed control module is a PID controller based on the BP neural network. On the basis of the classic PID controller, the measured speed returned by the Arduino module is compared with the theoretical speed in the algorithm processing system, and the BP neural network is used to Optimize the parameters of the PID controller, so as to realize the adaptive adjustment of the controller parameters of the robot in different motion states, and correct the speed deviation in real time.

The specific steps of the method are as follows:

Step 1. On the working platform, use the global positioning system to obtain the position coordinate information of the _{leader -} following _robot in the formation _relative to the global coordinate system _of the platform. ,θ _i-1 );

Step 2. According to the position coordinate information of each robot, construct the relative position angle relationship between the leader robot and the follower robot, define the distance angle relationship between the leader robot and the follower robot expected by the formation, obtain the error dynamics equation, and then derive the follower robot motor control rate;

Δx _i,i-1 = x _i -x _i-1 -dcosθ _i-1

Δy _i,i-1 = y _i -y _i-1 -dsinθ _i-1

Among them, Δx _i,i-1 and Δy _i,i-1 are the relative positional relationship between the leading robot and the following robot, and d is the distance from the front of the robot to the center of mass;

Define the desired distance-angle relationship of the lead-following robot in the formation in, They are the relative distance and angle respectively, and the error dynamic equation is obtained, and then the motion control rate of the following robot is deduced as:

where k ₁ k ₂ are variable parameters;

Step 3. During the formation movement, the leader robot moves autonomously to the target point and continuously sends information to the follower robot through the wireless communication system, and the follower robot receives the information of the leader robot and calculates its own position. Considering that the leader robot in the follower leader formation model It is easy to have abnormal communication data between the robot and the following robot. Record the location information of the following robot at each moment as training data, and establish a trajectory prediction model for the following robot based on the nonlinear least square method. The formula is as follows:

X(t,A)＝a ₀ +a ₁ t+a ₂ t ² +…a _n t ⁿ Among them, n=1,2…m are m sets of training data, A=(a ₀ ,a ₁ ,…, a _n ) is a set of parameters of the prediction model, and t is the time constant. We assume that the follower has normal communication data during the period of initial movement, and use the historical data of the follower during this period as the training set of the prediction model, and obtain the estimated value of A by training the parameters to be estimated That is, the predictive model is $\hat{x}(t,\hat{A})={\hat{a}}_0+{\hat{a}}_{1}t+{\hat{a}}_2{t}^2+...{\hat{a}}_{\mathrm{no}}{t}^{\mathrm{no}};$

Step 4, use the particle swarm algorithm to solve the problem of parameter selection of the trajectory prediction model, take A=(a ₀ ,a ₁ ,…,a _n ) as a set of solutions of the trajectory prediction model, initialize the particle swarm, according to the formula Calculate the fitness function of each particle, update the individual extremum of the particle according to the fitness function value of the particle, and use the optimal value of the individual extremum to update the global extremum; where, X _j and are the actual value and estimated value of the particle position at each moment, respectively;

Step 5, use the iterative formula to update the velocity and position of each particle; the formula is as follows:

v _id ＝ω×v _id +c ₁ ×rand()×(pbest _id -x _id )+c ₂ ×Rand()×(gbest _id -x _id )

x _id = x _id + v _id

Among them, the whole particle swarm contains N particles, the dimension of each particle is d _, v _id and x _id are the description of the speed and position of the i-th particle in the d dimension respectively, pbest is the individual extremum of the particle, and gbest is the global extremum value, c _{1 and} c ₂ are the learning factors, which are the weights of the degree of attraction of each particle to the global extremum and individual extremum positions, rand() and Rand() are random numbers in [0,1], ω is the inertia weight .

Among them, considering that when the system is in the initial stage of model establishment, the particle swarm needs to search for the optimal position at a faster speed to build a prediction model, and a finer local search is required in the later stage, so let

$\mathrm{<span\; class="notranslate">\; \&omega;</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span>\frac{\mathrm{<span\; class="notranslate">\; k</span>}}{{\mathrm{<span\; class="notranslate">\; k</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}}$

Among them, ω _max and ω _min are the maximum and minimum values of the inertia weight, k and k _max are the current iteration number and the maximum iteration number respectively, and ω shows a linear decreasing trend with the increase of the iteration number;

Step 6, continuously update the trajectory prediction model with the data at the next moment, and predict the trajectory information backwards, define an abnormal range of communication data,

like

Then it is determined that (x′ _i-1 , y′ _i-1 ) is an abnormal data point. Where ( _xi-1 , y _i-1 ) is the trajectory point following the robot at the current moment, (x′ _i-1 , y′ _i-1 ) is the trajectory point following the robot at the next moment, x _th and y _th are respectively It is the maximum threshold of the data exceeding the normal range on the horizontal and vertical coordinates, which is determined according to the global positioning and wireless communication system to collect each frame of image and calculate the time interval for sending the robot position information and the average moving speed of the robot. Therefore, the trajectory point of the following robot at the next moment is

$<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\left\{\begin{array}{cc}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>& \begin{array}{cc}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; f</span>}& <span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\end{array}\\ <span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>& \begin{array}{cc}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; f</span>}& <span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; r</span>}\end{array}\end{array}\right.$

Where (x _p+1 , y _p+1 ) is the estimated value of the prediction model. If the received data is within the normal range, calculate the position information of the following robot based on the received data of the leading robot and move towards it. If the data is abnormal, use the predicted point to replace the abnormal data point and calculate it as the follower at the next moment track point and move towards it;

The data update strategy is that the following robot trajectory point value at the next moment is determined according to the relationship between the following robot trajectory point value at the current moment and the maximum threshold value of the data exceeding the normal range, and by separately judging the current moment and the next moment The update decision is made by following whether the distance of the robot trajectory point value along the X and Y axis directions of the platform meets the abnormal range of the communication data.

Step 7, use the global positioning system to continuously update the position and attitude information of each robot in the formation until the leader follows the formation to reach the target point.

Fig. 2 is a flow chart of the control method of the method of the present invention. First initialize the position of each robot, initialize the particle swarm, initialize the formation system, and then start to move, the follower obtains the position information of the leader and calculates its own position information, at the same time records the training data, uses the particle swarm algorithm to optimize the prediction model parameters, and uses the optimized The last parameter to establish a prediction model, and then continuously update the prediction model with the data at the next moment, and predict backward to obtain a new estimated trajectory. Followers make judgments on their own position information obtained. If the data is within the normal range, Then the follower moves according to the given trajectory. If there is an abnormality in the short-term communication data, the estimated point is used as the follower trajectory point, and the leader follower moves according to the trajectory point to judge whether the formation meets the requirements. If not, then Each robot performs self-regulation. If it meets the requirements, it will judge whether it has reached the target point. If it does not reach the target point, it will perform the above operations in a loop. If it reaches the target point, the formation movement will end.

The above-mentioned embodiments are only descriptions of preferred implementations of the present invention, and are not intended to limit the scope of the present invention. All such modifications and improvements should fall within the scope of protection defined by the claims of the present invention.

Claims (8)

1. A control method for formation of a team of mobile robots based on piloting following is characterized by comprising a global positioning system, a wireless communication system, an algorithm processing system, a speed regulating system and mobile robots for calibrating different colors; the global positioning system utilizes the camera to carry out color identification on the mobile robots with different colors, and acquires the pose information of the mobile robots; the wireless communication system sends the collected pose information to each robot; the algorithm processing system performs motion control and coordinated formation algorithm processing according to the received position and attitude information of each mobile robot, continuously performs information interaction with the speed regulating system, the speed regulating system acquires and processes the speed information of each mobile robot, continuously performs feedback on actual control output, corrects speed deviation in real time, and finally completes the control of formation motion of each mobile robot.

2. The pilot following-based mobile robot formation control method according to claim 1, wherein the method comprises the following steps:

step 1, in a working platform, acquiring position coordinate information of each robot in a formation relative to a platform global coordinate system by using a global positioning system;

step 2, constructing a relative position angle relation between the piloting robot and the following robot according to the position coordinate information of each robot, defining a distance angle relation between the piloting robot and the following robot expected by formation, obtaining an error dynamics equation, and further deducing the motion control rate of the following robot;

step 3, in the process of formation movement, the piloting robot autonomously moves to a target point and continuously sends information to the following robot through a wireless communication system, the following robot receives the piloting robot information and calculates the position of the following robot, the position information of the following robot at each moment is recorded as training data, and a following robot track prediction model based on a nonlinear least square method is established;

step 4, solving the problem of selecting parameters of the track prediction model by using a particle swarm algorithm, taking a group of parameters of the track prediction model as a group of solutions of the track prediction model, initializing the particle swarm, calculating a fitness function of each particle, and updating an individual extreme value and a global extreme value of the particle according to the fitness function value of the particle;

step 5, updating the speed and the position of each particle by using an iterative formula;

step 6, continuously utilizing data at the next moment to update a track prediction model, predicting track information backwards, defining a communication data abnormal range, and calculating position information of the following robot according to the received data of the navigation robot and moving the following robot to the following robot if the received data is in the normal range; if the data is abnormal, starting the prediction points to replace abnormal data points as the following robot track points at the next moment;

and 7, continuously updating the position and posture information of each robot in the formation by using a global positioning system until the pilot following formation reaches a target point.

3. The pilot following-based mobile robot formation control method according to claim 1, wherein: the global positioning system is composed of an upper computer and a high-definition CCD camera, the high-definition CCD camera is arranged at the top of the working platform, different colors are marked for each mobile robot in the platform, position and posture information of the robot in the platform is obtained through a color recognition algorithm, the information is transmitted back to the upper computer for processing, and the information is transmitted to each mobile robot after being processed by the upper computer, so that the global positioning of the mobile robots is realized.

4. The pilot following-based mobile robot formation control method according to claim 1, wherein: the wireless communication system comprises an upper computer, a mobile robot and a wireless router, wherein a wireless transceiver module is installed on the mobile robot, the upper computer and the wireless router are connected through a network cable, the upper computer, the mobile robot and the wireless router form a communication system based on a local area network, and the wireless communication between the upper computer and the robots and between the robots is realized by establishing communication protocols between the upper computer and the robots and between the robots.

5. The pilot following-based mobile robot formation control method according to claim 1, wherein: the algorithm processing system consists of a DSP (digital signal processor) and is installed in the robot, a pilot following-based formation motion control algorithm of the mobile robot is implanted in the DSP, data from the wireless communication system is received, and autonomous motion and formation coordination control of the robot are processed.

6. The pilot following-based mobile robot formation control method according to claim 1, wherein: the speed regulating system comprises a speed measuring code disc module, an Arduino module and a PID speed regulating module;

the speed-measuring coded disc module consists of a grating disc and a photoelectric detection device, is arranged beside a left wheel motor and a right wheel motor of the robot, reflects the current revolution number of the left wheel motor and the right wheel motor by calculating the number of pulses output by a photoelectric encoder per second, and is converted by combining the circumference of a circle of wheel rotation to obtain the actual rotating speed of the left wheel and the right wheel of the motor of the mobile robot;

the Arduino module is an ArduinoMega2560 single chip microcomputer, is arranged in the robot and is used for expanding digital and analog input and output ports of the robot, further collecting and receiving the number of output pulses of a speed measuring code disc, analyzing and processing the data, and sending the data to an algorithm processing system through an I2C bus;

the PID speed regulation module is a PID controller based on a BP neural network, and compares the error of the actual measurement speed returned by the Arduino module with the theoretical speed in the algorithm processing system on the basis of the classical PID controller, and optimizes the PID controller parameters by using the BP neural network, thereby realizing the self-adaptive regulation of the controller parameters of the robot under different motion states and correcting the speed deviation in real time.

7. The pilot following-based mobile robot formation control method according to claim 2, wherein: in step 6, the communication data abnormal range is to use the global positioning system and the wireless communication system to collect each frame of image, and calculate the time interval for sending the position information of the robot and the average movement speed of the robot to determine whether the communication data is abnormal data.

8. The pilot following-based mobile robot formation control method according to claim 2, wherein: in step 6, the data update strategy is that the following robot track point value at the next moment is determined according to the maximum threshold relationship between the following robot track point value at the current moment and the data exceeding the normal range, and an update decision is made by respectively judging whether the distances between the following robot track point values at the current moment and the next moment along the axis direction of the platform X, Y meet the abnormal range of the communication data.