A vehicle tracking method and system based on fuzzy adaptive control

Abstract

The present invention provides a vehicle tracking method and system based on fuzzy adaptive control, comprising the following steps: step 1, preprocessing a tracking image to obtain a preprocessed tracking image; step 2, edge extraction of the preprocessed tracking image to obtain a tracking band; step 3, obtaining a distance deviation and an angle deviation between a robot and the tracking image according to the obtained tracking band; step 4, using the obtained distance deviation and angle deviation as input quantities of a fuzzy adaptive control algorithm to obtain a vehicle tracking result; the method meets the real-time requirements, the system corrects the running state of the vehicle in real time, and the effect is obvious.

Description

Technical Field

[0001] The present invention belongs to the field of robots, and in particular relates to a vehicle tracking method and system based on fuzzy adaptive control. Background Art

[0002] The study of tracking control methods is an important step in achieving line-following driving. The rationality of tracking control methods is directly related to the stability and accuracy of tracking driving. At present, the methods used for tracking control mainly include classic PID control methods, fuzzy control methods and neural network control methods. These control methods have their own advantages and disadvantages. Although control algorithms such as neural networks are relatively advanced, they are not mature enough in the field of motion control. Therefore, the commonly used control methods for motion control are PID control and fuzzy control. PID is a classic control method with high reliability, good robustness, relatively simple algorithm, and wide application. It is most widely used in engineering application fields, but it requires a relatively accurate system model, the control parameter setting is complex, and it is generally used in single-input and single-output control occasions. The fuzzy control method does not require the establishment of an accurate mathematical model, has strong adaptability, simulates the processing method of the human brain, and can meet the requirements of the motion control system with the help of experience, and can play its advantages in the motion control system. Summary of the invention

[0003] The purpose of the present invention is to provide a vehicle tracking method and system based on fuzzy adaptive control, which solves the above-mentioned deficiencies in the prior art.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is:

[0005] The present invention provides a vehicle tracking method based on fuzzy adaptive control, comprising the following steps:

[0006] Step 1, preprocessing the tracking image to obtain a preprocessed tracking image;

[0007] Step 2, extracting the edge of the preprocessed tracking image to obtain a tracking band;

[0008] Step 3, obtaining the distance deviation and angle deviation between the robot and the tracking image according to the obtained tracking band;

[0009] Step 4: Use the obtained distance deviation and angle deviation as input of the fuzzy adaptive control algorithm to obtain the vehicle tracking result.

[0010] Preferably, in step 4, the fuzzy adaptive control algorithm includes a basic fuzzy controller, and the basic fuzzy controller includes a deviation measurement module, a control correction module and a rule modification module, wherein:

[0011] The deviation measurement module is used to calculate the deviation measurement and correction amount corresponding to the distance deviation and the angle deviation respectively;

[0012] The control correction module is used to convert the result output by the deviation measurement module into an input correction amount of the controlled object and apply it to the controlled object;

[0013] The rule modification module is used to modify the control rule of the basic fuzzy controller according to the obtained input correction amount of the controlled object.

[0014] Preferably, the control rule of the basic fuzzy controller is modified according to the obtained input correction amount of the controlled object, and the specific method is:

[0015] Assuming the time lag of the controlled object is h beats, the time when the system's poor performance is discovered at the current time k is (kh);

[0016] Obtain the position deviation, angle deviation and control amount corresponding to the time (kh);

[0017] The obtained position deviation, angle deviation and control quantity are fuzzy processed to obtain new fuzzy control rules;

[0018] The obtained fuzzy control rules are added to the basic fuzzy controller.

[0019] Preferably, in step 4, the distance deviation and angle deviation between the robot and the tracking image are used as inputs of the fuzzy adaptive control algorithm to obtain the vehicle tracking result. The specific method is:

[0020] The obtained input variables are fuzzified;

[0021] Establish fuzzy adaptive control rules;

[0022] Defuzzification processing is performed to obtain the vehicle tracking result.

[0023] Preferably, in step 1, the tracking image is preprocessed to obtain a preprocessed tracking image, and the specific method is:

[0024] The tracking image is subjected to grayscale processing, filtering processing, binarization processing and morphological processing in sequence to obtain a preprocessed tracking image.

[0025] Preferably, in step 3, the distance deviation and angle deviation between the robot and the tracking image are obtained according to the obtained tracking band, and the specific method is:

[0026] Extract the center line from the obtained tracking band to obtain the coordinates of the center point of each row of the tracking band;

[0027] Fit the obtained midline point coordinates to obtain a fitting straight line;

[0028] The distance deviation and angle deviation between the robot and the tracking image are calculated based on the obtained fitting straight line.

[0029] A vehicle tracking system based on fuzzy adaptive control, comprising:

[0030] A preprocessing module, used for preprocessing the tracking image to obtain a preprocessed tracking image;

[0031] An edge extraction module is used to extract the edge of the preprocessed tracking image to obtain a tracking band;

[0032] A calculation module, used for obtaining a distance deviation and an angle deviation between the robot and the tracking image according to the obtained tracking belt;

[0033] The output module is used to use the obtained distance deviation and angle deviation as the input of the fuzzy adaptive control algorithm to obtain the vehicle tracking result.

[0034] A vehicle tracking device based on fuzzy adaptive control comprises a processor and a computer program that can be run on the processor. When the processor executes the computer program, the steps of the method are implemented.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] The present invention provides a vehicle tracking method based on fuzzy adaptive control, which obtains path image information through visual detection, and obtains position deviation and angle deviation through digital image technology processing. The two deviations are used as inputs of a fuzzy controller, and clear and definite outputs are obtained through fuzzy operation processing to control the driving posture of the vehicle, so that it runs smoothly along the tracking belt. After image preprocessing operations, interference is reduced and the operation efficiency is greatly improved. The adaptive control method designed in this application is added to the fuzzy control method to form an adaptive fuzzy control method, and vehicle tracking is completed through the adaptive fuzzy control method. The method meets the real-time requirements, and the system corrects the running state of the vehicle in real time, with obvious effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of the process of the present invention;

[0038] Figure 2 is the triangle membership function graph;

[0039] Figure 3 It is the fuzzy control distance deviation variation curve;

[0040] Figure 4 is the fuzzy control angle deviation variation curve;

[0041] Figure 5 It is the fuzzy adaptive control distance deviation variation curve;

[0042] Figure 6 It is the fuzzy adaptive control angle deviation change curve. DETAILED DESCRIPTION

[0043] The invention will be further described in detail below in conjunction with the accompanying drawings.

[0044] like Figures 1 to 3 As shown, the present invention provides a vehicle tracking method based on fuzzy adaptive control, comprising the following steps:

[0045] Step 1: Preprocess the tracking image to obtain the preprocessed tracking image

[0046] (1) Grayscale processing of tracking images

[0047] Grayscale processing methods mainly include maximum value method, average value method and weighted average method. This patent adopts the average value method for grayscale processing. The average value method is to average the brightness of the three components in the color image to obtain a grayscale value. The formula of the average value method is shown in the following formula (1):

[0048]

[0049] (2) Tracking image filtering processing

[0050] The filtering processing methods mainly include median filtering, mean filtering and Gaussian filtering. This patent adopts median filtering for filtering. Median filtering is a nonlinear filtering method. It uses a convolution kernel to perform a convolution operation on the original image, calculates the median value of the template coverage area, and replaces the pixel value at the center of the template with the result. When the convolution kernel scans the image, the median filtering operation is completed.

[0051] (3) Binarization of tracking images

[0052] There are many binarization methods, and threshold segmentation is one of the classic binarization methods. It divides pixels into several categories by setting thresholds based on the difference in grayscale between the target area and the background in the image, thereby achieving separation of the target area and the background. The purpose of threshold segmentation is to convert a grayscale image into a binary image. It compares the value of each pixel in the image with the set threshold to determine whether the pixel belongs to the target area or the background area. The characteristics of the threshold segmentation method are simple calculation and the ability to always define non-overlapping areas with closed and connected boundaries. The threshold segmentation formula is shown in the following formula (2):

[0053]

[0054] Among them, f(x, y) is the original image and T is the threshold.

[0055] The binarization method used in this patent is the maximum inter-class variance method in the threshold segmentation method, also known as the Otsu method, which is a method for automatically determining the threshold based on the maximum inter-class variance. It is a binarization algorithm based on global considerations. By virtue of the grayscale characteristics of the image, the image can be divided into two parts, namely the foreground and the background. When the optimal threshold is taken, the difference between the two parts is the largest. The larger the inter-class variance between the foreground and the background, the greater the difference between the two parts that constitute the image. Therefore, the segmentation with the largest inter-class variance means the smallest probability of misclassification. The formula of the maximum inter-class variance method is shown in the following formula (3):

[0056] g=ω0·ω1·(u0-u1) 2 (3)

[0057] Among them, ω0 is the ratio of the number of foreground points to the image, ω1 is the ratio of the number of background points to the image, u0 is the average grayscale of the number of foreground points, and u1 is the average grayscale of the number of background points.

[0058] When the variance g is the largest, the difference between the foreground and the background is the largest, and T is the optimal threshold.

[0059] (4) Morphological processing of tracking images

[0060] Long-term tracking of the robot may cause the tracking belt to be damaged, or other reasons may make the tracking belt incomplete. The purpose of morphological processing is to restore the damaged tracking belt extracted by vision so that the robot tracking is not affected.

[0061] Morphological processing methods mainly include corrosion, expansion, opening and closing operations. This patent uses closing operations for morphological processing. Closing operation is the process of first performing an expansion operation on the image to be processed and then performing an corrosion operation. This method can be used to eliminate small black holes. The closing operation will eliminate isolated points with relatively low pixel values. For smaller debris and small areas of damage in the image, the closing operation can achieve the desired effect. First, the expansion allows the damaged points of the tracking track to be completed and repaired, and then the corrosion reduces the tracking track to the original effect.

[0062] Step 2, perform edge extraction on the preprocessed tracking image obtained in step 1 to obtain a tracking band, that is, the tracks on both sides of the tracking band. The track and the part of the tracking band surrounded by the track can be extracted from the image, that is, the blank area in the image after binarization. A straight line fitting is performed on the basis of the extracted tracking band to obtain the input of the fuzzy control.

[0063] Edge extraction operators mainly include Roberts operator, Sobel operator, Prewitt operator, LoG operator and Canny operator, etc. This patent uses Canny operator for edge extraction. The Canny operator first performs Gaussian filtering to denoise the image, then uses the Sobel operator (the Canny operator contains the Sobel operator) to calculate the gradient, and finally uses the non-maximum condition to remove the non-edge part and retain the edge information. The advantage is that two different thresholds are used to detect strong edges and weak edges respectively, and when the weak edge is connected to the strong edge, the weak edge is included in the output image. The Canny operator is not easily affected by noise and can detect real weak edges. The Canny edge detection operator not only has a fast detection speed, but also has a relatively high accuracy.

[0064] Step 3, extract the center line of the path tracking tape obtained in step 2 to obtain the coordinates of the center line points of each row of the path tracking tape; then fit a straight line; and obtain the distance deviation d and the angle deviation θ between the robot and the tracking image based on the fitted straight line.

[0065] After a series of processing on the tracking image, two edge lines of the tracking image have been obtained. However, the final result of the processing is to obtain the distance deviation d and angle deviation θ between the robot and the tracking image. Therefore, it is necessary to extract the tracking belt from the tracking image, and then calculate d and θ based on the shape of the tracking belt and its geometric relationship.

[0066] This application uses the statistical method of pixel points to extract the tracking band. Because in the tracking image, the tracking band occupies a certain area, so the contour area with the largest pixel points is the required tracking band. Based on this method, the tracking band can be extracted more ideally. After the tracking band is extracted, it needs to be further processed.

[0067] Fitting a straight line reflects the path information. In order to obtain the fitted straight line, the least square method is used to fit the straight line with the help of the physical coordinate system, and the fitted straight line in the physical coordinate system is set to: y=ax+b.

[0068] The average value of the pixel coordinates of each row in the tracking band is calculated to obtain 200 sets of data (x i ,y i )i=1,2,3,……,200,observation value y i The weighted sum of squares of the deviations should be minimized according to the requirements of the least squares method, as shown in the following formula (4):

[0069]

[0070] Among them, x iis the average value of the horizontal coordinates of the pixel coordinates from left to right in the i-th row, a is the intercept of the fitting line, and b is the slope of the fitting line.

[0071] The above formula is used to obtain partial derivatives of α and b respectively and then rearranged to obtain:

[0072]

[0073]

[0074] According to the geometric relationship, let y = 0 to obtain d = _b / a, θ = arctan (1 / a), d is the position deviation between the robot and the tracking belt, θ is the angle deviation between the robot and the tracking belt, d and θ are used as the input of fuzzy control.

[0075] Step 4: Fuzzy control algorithm tracking

[0076] First, the two input variables obtained after processing in step 3 are fuzzified, that is, the distance deviation d and the angle deviation θ are fuzzified to obtain the distance deviation and angle deviation of the robot relative to the tracking belt, which are used as the input of fuzzy control;

[0077] Then, the control rules are determined according to the operation requirements, and fuzzy reasoning is performed;

[0078] Finally, the inference results are clarified and a control signal is generated to make the robot run along the tracking belt by controlling the speed difference between the two driving wheels.

[0079] In the control system, the input of the fuzzy controller is the position deviation and angle deviation of the robot body relative to the path tracking belt. This information is obtained by processing the path image obtained by the camera.

[0080] (1) Fuzzification

[0081] The fuzzification of the input and output control variables of the fuzzy control system is to describe the position deviation and angle deviation of the robot relative to the path tracking belt obtained by image processing using fuzzy quantization factors, and then use the control rules for fuzzy reasoning. Usually, the determined system variables are discretized and divided into different levels. The set corresponding to each level is a fuzzy subset. In a fuzzy subset set, its members can be described by a fuzzy language. This process is the fuzzification process of the system variables. For the control system in this study, the fuzzification process can be divided into the following two steps:

[0082] Fuzzy processing is simulated using MATLAB. The command can be called in the command window to design the fuzzy controller. After the fuzzy input and fuzzy output and fuzzy rules are manually designed, MATLAB will perform fuzzy reasoning autonomously according to Mamdan i reasoning and output the results.

[0083] The first step is to transform the precise input variables and output variables obtained by the control system into the corresponding fuzzy set theory area. The original range of precise variables is called the basic domain.

[0084] In order to ensure the stability of the robot during operation, the position deviation range and output variable range need to be limited. The position deviation range is [-0.05, 0.05] (unit: m), and the angle deviation range is [-30°, 30°]. At the same time, the output of the fuzzy system is also limited. The range of ΔV is [-0.2, 0.2] (unit: m), and ΔV is the wheel speed difference of the robot's driving wheel. It is stipulated that the position deviation of the robot on the right side of the tracking belt is positive, and the angle deviation in the reverse and forward directions relative to the tracking belt is positive. According to the robot control requirements, the fuzzy set domain of the fuzzy system is set to [-6, 6], and the fuzzy language values ​​are also set to NB (negative large), NM (negative medium), NS (negative small), Z0 (zero), PS (positive small), PM (positive medium), PB (positive large).

[0085] In the second step, after determining the fuzzy level of the variables, it is necessary to specify the membership function of each subset.

[0086] The role of membership is to evaluate the function. For a system with an uncertain mathematical model, its output cannot be simply judged by yes or no, but needs to be described by fuzzy variables.

[0087] For an element x in the domain U, a mapping relationship is formed with the function set A(x). This correspondence is definite and one-to-one. At the same time, the value range of A(x) is specified as [0, 1]. Then A(x) is called the fuzzy set of x on the domain U. The membership of x to A is the function A(x), that is, A(x) is the membership function of x. This application uses the triangular membership function to perform fuzzification processing of fuzzy input.

[0088] The triangle membership function is determined by three parameters a, b, and c. The mathematical expression is shown in equation (7) and the image is shown as follows: Figure 2 As shown:

[0089]

[0090] Among them, parameters a and c determine the foot of the triangle, and parameter b determines the peak of the triangle.

[0091] (2) Establishing fuzzy control rules

[0092] After the control variables are determined, the next step is to write the control rules based on experience. For fuzzy control systems, the key link is the fuzzy reasoning process based on control rules, and this application adopts Mamdani reasoning. The more fuzzy control rules are determined, the more detailed the actual situation is, so that the fuzzy control system can adapt to different application requirements. However, this will also cause new problems, that is, the processing time of the control system will become longer, so it is necessary to reasonably determine the fuzzy control rules and the number of rules in the fuzzy control system. By consulting relevant information on fuzzy control and experimental simulation experience, the control rules for this article are determined, and these control rules are converted into corresponding control languages.

[0093] In this application, the input variables of the control system are the angle deviation θ and the position deviation d. The purpose of fuzzy control is to obtain the speed difference of the two driving wheels of the control output by fuzzy processing the two input variables, and correct the running state of the robot to make it run smoothly. The fuzzy control rule table is shown in Table 1 below:

[0094] Table 1 Fuzzy control rules table

[0095]

[0096]

[0097] Add the adaptive module to form fuzzy adaptive control:

[0098] Generally speaking, control rules are summarized and designed artificially. However, if the controlled object or the external environment is complex, not only will the difficulty of compiling and adjusting the control rules increase, but they will also be unable to adapt to the dynamic and static mutations in actual operations, and the corresponding control effect will also be affected.

[0099] Therefore, this application designs a fuzzy adaptive control method to solve this problem. This method does not rely on manual methods, but grasps relevant information from the external environment and its own control process, and identifies, calculates, and clusters it based on this, and automatically modifies the control rules or parameters, so as to make the dynamic and static performance of the system more perfect.

[0100] The fuzzy adaptive control method in this application adds three functional modules to the basic fuzzy controller: deviation measurement, control correction and rule modification. The structure applied to the control system is as follows Figure 1 shown.

[0101] ① Deviation measurement and correction amount acquisition

[0102] For the input variables d and θ, let the current time be k, and obtain the sampling value distance deviation dk and angle deviation θ by measurement. k, whose size reflects the deviation from the expected value at the current moment, so that the amount x(k) required for correction can be calculated to reduce the deviation until it reaches zero.

[0103] By considering the actual situation, it is more appropriate to determine the deviation measurement and correction amount after fuzzification, that is, to determine the exact amount from the real domain to the domain through statistical methods, d(k) from the real domain [-0.05, 0.05] to the domain [-6, 6], and θ(k) from the real domain [-30°, 30°] to the domain [-6, 6].

[0104] Assume that the domain of the correction amount is [-3, 3]. According to the same idea as the basic controller control table, the level of the correction amount x(k) is determined according to the size and positive and negative of d(k) and θ(k). The deviation measurement correction table is shown in the following table:

[0105]

[0106] The rows in the table are the levels of change in the deviation d, the columns are the levels of change in the deviation θ, and the elements in the table are the levels of change in the correction amount x(k).

[0107] ②Conversion of correction amount to control input correction amount

[0108] The correction amount x(k) obtained by the deviation measurement needs to be converted into the input correction amount of the controlled object, that is, the control correction amount.

[0109] And applied to the controlled object, the deviation can be reduced and the output can reach the desired value. Therefore, it is necessary to determine the input and output relationship of the controlled object; regarding the input and output relationship of the controlled object, it is best to establish a relevant model. Considering that the controller of this application has a dual-input and single-output structure, specifically:

[0110] Let u = ΔV, then the state space expression is u = f(d, θ).

[0111] When the input changes slightly, the output changes to

[0112] In a sufficiently short sampling period T, the change in output Δu caused by the change in input Δd and Δθ can be approximately expressed as

[0113]

[0114] In the formula, M is the output coefficient, so it is known that the correction amount x(k) is obtained by deviation measurement, and after conversion, the input correction amount u of the controlled object is obtained c (k) = M -1 x(k).

[0115] ③Rule modification

[0116] After obtaining the control correction, the next step is to write it into the control rules to improve the control performance. Since the fuzzy controller has control rules, the addition of new control correction is actually to modify the control rules.

[0117] Assuming that the time lag of the controlled object is h beats, i.e. hT, the poor performance of the system at time k should be the situation at time (kh), when the position deviation d, angle deviation θ and uncorrected fuzzy output are d(kh), θ(kh) and u(kh) respectively. After taking into account the correction amount, the current control amount should be u(kh)+u c (k) In order to modify the control rules and implement corrections, these quantities need to be fuzzy processed.

[0118] Based on this, the original fuzzy control rule is: if d is d i andθisθ i Then ΔV is u i ;

[0119] Then the current fuzzy control rule should be: if d is d i (kh)andθisθ i (kh)thenΔVis u i (kh).

[0120] Add the modified control rules to the system and then perform the subsequent processing to complete the real-time fuzzy adaptive control.

[0121] (3) Defuzzification: The actual deviation d and θ between the robot and the tracking line are obtained by straight line fitting as the input of fuzzy control. The input is then fuzzified and fuzzy rule reasoning is performed to obtain the fuzzy output. Defuzzification is to transform the fuzzy output into the actual speed difference between the two driving wheels of the robot. After obtaining the true value, the two driving wheels are controlled and adjusted to complete the tracking function.

[0122] The main purpose of defuzzification is to map the fuzzy output value obtained by the fuzzy control system to the real world, instead of expressing it in the form of fuzzy language. The control output obtained by fuzzy reasoning based on certain fuzzy rules is also a fuzzy variable, but actual control requires a certain control quantity. The fuzzy quantity output of the fuzzy control needs to be processed to obtain the determined output quantity required for actual control. This process is called clarification of fuzzy quantity, also known as declarative.

[0123] This application uses the area center method (i.e., the center of gravity method) to defuzzify, and uses the median in the fuzzy control system as the control quantity. The center of gravity method contains all the information in the fuzzy subset, and the output is smoother. The area center method is to find the center of the area of ​​the area surrounded by the fuzzy set membership function curve and the horizontal coordinate, and select the horizontal coordinate value corresponding to this center as the representative value of this fuzzy set, which is equivalent to treating the area as the center of gravity of a plate of equal thickness.

[0124] For the discrete domain U(x1, x2, ..., x n ), x i The degree of membership at is A(x i ), then the horizontal coordinate x corresponding to the center of the area cen It can be calculated by the following formula (8):

[0125]

[0126] The image processing part extracts the contour of the tracking band, and then fits the straight line to obtain the tracking input, i.e., d and θ. After fuzzifying the input, fuzzy reasoning is performed, and an adaptive adjustment mechanism is added to obtain the fuzzy output. After defuzzification, the true output, i.e., ΔV, is obtained.

[0127] By building a fuzzy controller simulation model in the Simulink module of Matlab simulation software, setting the angle deviation θ = 30°, the distance deviation d = 0.1m, and the initial speed of the left and right wheels to 0.4m / s, the simulation diagram of the fuzzy controller is finally obtained as shown below: Figures 3 to 6 The time response curve of distance deviation d and the time response curve of angle deviation θ under the fuzzy controller are shown.

[0128] Among them, from Figure 3 and Figure 4 From the analysis of the time response curve of the fuzzy control distance deviation d and the time response curve of the fuzzy control angle deviation θ shown in the figure, it can be seen that: when the initial distance deviation and angle deviation are given, that is, when the vehicle trajectory deviates, the fuzzy controller can respond in time, and the angle deviation θ will gradually decrease immediately, but the distance deviation d cannot be reduced immediately. As the vehicle running time slowly advances, the distance deviation d gradually begins to decrease, while the angle deviation increases in the opposite direction due to inertia.

[0129] In the simulation diagram, it can be seen that the angle deviation and distance deviation of the vehicle using the fuzzy controller will tend to zero in a very short time, and the vehicle trajectory can return to the predetermined trajectory position, achieving the purpose of trajectory tracking. By adding adaptive control links to fuzzy control, a more ideal graph can be obtained. Figure 5 and Figure 6 As shown, the distance deviation and the angle deviation fluctuate less, and the deviation recovery time can be further improved.

[0130] Summarizing the above analysis of the simulation diagram, the trajectory tracking controller using the fuzzy adaptive control principle takes less time to eliminate the deviation under the same deviation and the number of fluctuations of the deviation approaching zero is smaller. Therefore, it can be clearly seen that the vehicle trajectory tracking controller based on the fuzzy adaptive control principle has stronger system stability and better deviation elimination effect than the fuzzy control principle controller, and can be better applied to the actual operation of the vehicle.

[0131] In another embodiment of the present invention, a vehicle tracking system based on fuzzy adaptive control is provided, which can be used to implement the above-mentioned vehicle tracking method based on fuzzy adaptive control. Specifically, the vehicle tracking system based on fuzzy adaptive control includes a preprocessing module, an edge module, an extraction module, a fitting module and an output module.

[0132] Wherein, the preprocessing module preprocesses the tracking image to obtain a preprocessed tracking image;

[0133] An edge extraction module is used to extract the edge of the preprocessed tracking image to obtain a tracking band;

[0134] A calculation module, used for obtaining a distance deviation and an angle deviation between the robot and the tracking image according to the obtained tracking belt;

[0135] The output module is used to use the obtained distance deviation and angle deviation as the input of the fuzzy adaptive control algorithm to obtain the vehicle tracking result.

[0136] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention described and shown in the drawings here can usually be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0137] In summary, the present invention is a vehicle tracking method and system based on fuzzy adaptive control. Please summarize and explain the effects of the present invention. By comparing the fuzzy adaptive control simulation results, it can be clearly seen that the fuzzy adaptive controller eliminates the deviation of the robot very well. It shows that the fuzzy adaptive controller is more suitable for the operation of the robot in a complex environment and the original intention of using fuzzy adaptive control design for the robot is effective, which makes the robot run more stable and safer.

[0138] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0139] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0140] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0141] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0142] The above contents are only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A vehicle tracking method based on fuzzy adaptive control, characterized in that: The following steps are involved: Step 1, preprocessing the tracking image to obtain a preprocessed tracking image; Step 2, extracting the edge of the preprocessed tracking image to obtain a tracking band; Step 3, obtaining the distance deviation and angle deviation between the robot and the tracking image according to the obtained tracking band; Step 4, using the obtained distance deviation and angle deviation as inputs of the fuzzy adaptive control algorithm to obtain the vehicle tracking result; In step 4, the fuzzy adaptive control algorithm includes a basic fuzzy controller, and the basic fuzzy controller includes a deviation measurement module, a control correction module and a rule modification module, wherein: The deviation measurement module is used to calculate the deviation measurement and correction amount corresponding to the distance deviation and the angle deviation respectively; The control correction module is used to convert the result output by the deviation measurement module into an input correction amount of the controlled object and apply it to the controlled object; The rule modification module is used to modify the control rule of the basic fuzzy controller according to the input correction amount of the controlled object; The control rules of the basic fuzzy controller are modified according to the input correction amount of the controlled object. The specific method is: If the time lag of the controlled object is set to h beats, then the moment when the system's poor performance is discovered at the current time k is kh; Obtain the position deviation, angle deviation and control amount corresponding to the moment kh; The obtained position deviation, angle deviation and control quantity are fuzzy processed to obtain new fuzzy control rules; Add the obtained fuzzy control rules to the basic fuzzy controller; In step 3, the distance deviation and angle deviation between the robot and the tracking image are obtained according to the obtained tracking band. The specific method is: Extract the center line from the obtained tracking band to obtain the coordinates of the center point of each row of the tracking band; Fit the obtained midline point coordinates to obtain a fitting straight line; The distance deviation and angle deviation between the robot and the tracking image are calculated based on the obtained fitting straight line.

2. The vehicle tracking method based on fuzzy adaptive control according to claim 1, characterized in that: In step 4, the distance deviation and angle deviation between the robot and the tracking image are used as the input of the fuzzy adaptive control algorithm to obtain the vehicle tracking result. The specific method is: The obtained input variables are fuzzified; Establish fuzzy adaptive control rules; Defuzzification processing is performed to obtain the vehicle tracking result.

3. The vehicle tracking method based on fuzzy adaptive control according to claim 1, characterized in that: In step 1, the tracking image is preprocessed to obtain a preprocessed tracking image. The specific method is: The tracking image is subjected to grayscale processing, filtering processing, binarization processing and morphological processing in sequence to obtain a preprocessed tracking image.

4. A vehicle tracking system based on fuzzy adaptive control, characterized in that: The method according to any one of claims 1 to 3, comprising: A preprocessing module, used for preprocessing the tracking image to obtain a preprocessed tracking image; An edge extraction module is used to extract the edge of the preprocessed tracking image to obtain a tracking band; A calculation module, used for obtaining a distance deviation and an angle deviation between the robot and the tracking image according to the obtained tracking belt; The output module is used to use the obtained distance deviation and angle deviation as the input of the fuzzy adaptive control algorithm to obtain the vehicle tracking result.

5. A vehicle tracking device based on fuzzy adaptive control, comprising a processor and a computer program that can run on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 3 are implemented.

Images