Robot control system based on big data and artificial intelligence

By using hybrid encoding and genetic algorithms to optimize the membership functions and rules of fuzzy PID controllers in robot control systems based on big data and artificial intelligence, the problems of insufficient rules and conflicts in PID controllers in the existing technology are solved, and the problems of improvement of the rules and regulations of PID controllers have been improved. Control effect and anti-interference ability reduce tracking deviation.

CN119017390AInactive Publication Date: 2024-11-26何楚翘
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411342186.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-25
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

In robot control systems based on big data and artificial intelligence, PID controllers may have problems such as insufficient number of rules, conflicts of rules or inconsistent with the actual situation when used, resulting in a decrease in control effect and an increase in tracking deviation, affecting anti-interference ability.

Method used

The hybrid encoding method is used to encode the design parameters and feature parameters of the fuzzy PID controller, and the genetic algorithm is used to optimize the membership function and fuzzy rules of the fuzzy PID controller. By generating the initial population, updating the population, updating the membership function and fuzzy rules, and calculating the The objective function value, determining whether to traverse individual chromosomes and determining whether to terminate the genetic optimization algorithm, etc., optimize the fuzzy PID controller.

Benefits of technology

减少了控制效果下降的概率,降低了跟踪偏差,并降低了对抗干扰能力的影响,提高了机器人控制的精确性和稳定性。

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119017390A_ABST
    Figure CN119017390A_ABST
Patent Text Reader

Abstract

The invention provides a robot control system based on big data and artificial intelligence, which relates to the technical field of robot control and comprises a data acquisition module, a data preprocessing module, a data storage module, a data analysis and intelligent processing module, a control algorithm optimization module, a control decision module and an execution module. According to the invention, the genetic algorithm is adopted to optimize the membership function and the fuzzy rule of the fuzzy PID controller; and determining a hybrid coding genetic algorithm optimization robot trajectory tracking control fuzzy PID controller through the following steps of generating an initial population, updating the population, updating a membership function and a fuzzy rule, calculating a target function value, judging whether to traverse N individual chromosomes, judging whether to terminate a genetic optimization algorithm and generating a next generation subgroup through genetic operation. A fuzzy PID controller is optimized through a hybrid coding method genetic algorithm, so that the probability that the control effect is reduced is reduced, the tracking deviation is reduced, and the influence on the anti-interference capability of the fuzzy PID controller is reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of robot control technology, and in particular to a robot control system based on big data and artificial intelligence. Background Art

[0002] Robots based on big data and artificial intelligence are robots that use advanced data processing technology and intelligent algorithms to enhance their functions and performance. These robots are able to process and analyze large amounts of data to achieve higher levels of automation, decision support, and interaction capabilities. They play an increasingly important role in multiple fields.

[0003] At present, robots based on big data and artificial intelligence will control robots through fuzzy PID controllers. First, the input of the robot control system is fuzzified and converted into membership values ​​in fuzzy sets, and a fuzzy rule base is established based on expert knowledge or experimental data. These rules describe the fuzzy relationship between input and output. Fuzzy reasoning is performed based on the input of the current system and the rules in the fuzzy rule base to calculate the fuzzy value of the output. The fuzzy reasoning process usually uses Mamdani reasoning or Takagi-Sugeno reasoning, and then the fuzzy value of the output is defuzzified and converted into a specific control quantity. Finally, according to the control quantity obtained by defuzzification, the proportional coefficient, integral coefficient and differential coefficient of the PID controller are adjusted in real time. However, when the PID controller is used, there may be insufficient rules, conflicts between rules, or the rules are inconsistent with the actual situation, which leads to a decrease in control effect, increases tracking deviation, and affects its own anti-interference ability.

[0004] Therefore, it is necessary to provide a new robot control system based on big data and artificial intelligence to solve the above technical problems. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a robot control system based on big data and artificial intelligence.

[0006] A robot control system based on big data and artificial intelligence provided by the present invention includes a data acquisition module, which is responsible for collecting various data from the robot's internal and external environment;

[0007] A data preprocessing module is used to perform preprocessing operations on the collected raw data;

[0008] A data storage module, used for storing the preprocessed data in a large-scale data storage system;

[0009] Data analysis and intelligent processing module, which uses big data processing technology and artificial intelligence algorithms to analyze and mine stored data;

[0010] The control algorithm optimization module adopts a hybrid coding method to encode the design parameters and characteristic parameters of the fuzzy PID controller, and uses a genetic algorithm to optimize the membership function and fuzzy rules of the fuzzy PID controller. At the same time, the individuals are genetically operated according to the fitness function corresponding to the selected objective function, and when running the genetic algorithm, each design parameter is directly represented by the individual chromosome gene position. Direct coding method;

[0011] The control decision module is used to formulate the robot's control strategy and instructions based on the output results of the data analysis and intelligent processing module;

[0012] The execution module is responsible for converting the instructions given by the control decision module into actual actions of the robot.

[0013] Preferably, optimizing the robot trajectory tracking control fuzzy PID controller by a hybrid coding-based genetic algorithm comprises the following steps:

[0014] S1. Generate the initial population. According to the range of the characteristic parameters and design parameters of the fuzzy controller, adopt the mixed encoding method of characteristic parameters and design parameters, and randomly generate N individual chromosomes [a1, a2, a3, ..., a 63 ];

[0015] S2, update the population, update the current population to the parent population;

[0016] S3, update the membership function and fuzzy rules, read the gene codes of the jth chromosome in the current population, and calculate the gene codes according to the codes a1-a 10 Generate new input variables e, ec and output variable ΔK P , ΔK I , ΔK D Membership function, update fuzzy rules;

[0017] S4, calculate the objective function value, run the fuzzy PID controller to complete the trajectory tracking control simulation, and calculate the deviation e(t) between the simulated trajectory of the robot and the expected trajectory at time t, and calculate the objective function value J;

[0018] S5, determine whether to traverse N individual chromosomes. If the number of individual chromosomes j does not satisfy j>N, then j+1 and enter S3; if j>N, then enter S6;

[0019] S6, judging whether to terminate the genetic optimization algorithm, judging whether the change of the objective function value is less than the set threshold or whether the genetic generation reaches the set value, if satisfied, then terminate the genetic optimization algorithm, if not satisfied, then enter S7;

[0020] S7, Genetic operations generate the next generation of sub-populations: Based on the fitness function value converted from the objective function value of each individual in the current population, a new population can be obtained by performing basic genetic algorithm operations such as selection, replication, crossover and mutation on all individuals in the population, and then enter S2.

[0021] Preferably, the specific principle of the fuzzy PID controller is: first, the deviation e and the deviation change ec of the controlled variable are used as the two input variables of the fuzzy controller, and after fuzzification, the membership degrees of e and ec belonging to different fuzzy sets are obtained, and then fuzzy reasoning is performed based on fuzzy rules to obtain the output variable ΔK P , ΔK I , ΔK D The fuzzy result is defuzzified and the precise result of the output variable is obtained.

[0022] Preferably, the design of a genetic algorithm to solve an optimization problem comprises the following steps:

[0023] S10, determining the parameters to be optimized and the optimization constraints such as the upper and lower limits of the parameters according to the optimization problem;

[0024] S20, designing a chromosome encoding representation method to achieve the mutual conversion relationship between the parameters to be optimized and the chromosome gene positions;

[0025] S30, determining a calculation method of an objective function and a fitness function of a genetic algorithm according to the target to be optimized;

[0026] S40, selecting basic operations of the genetic algorithm, and determining specific implementation methods of selection, replication, crossover and mutation;

[0027] S50, setting necessary parameters in the specific implementation method of the genetic operation in S40, such as crossover probability and mutation probability, and setting other genetic operation parameters such as population size and optimization termination condition.

[0028] Preferably, the various data include sensor data, environmental parameters and user input data.

[0029] Compared with the related art, the robot control system based on big data and artificial intelligence provided by the present invention has the following beneficial effects:

[0030] The present invention combines the design parameters and characteristic parameters of the fuzzy PID controller in a mixed coding manner, adopts a genetic algorithm to optimize the membership function and fuzzy rules of the fuzzy PID controller, and determines the mixed coding genetic algorithm to optimize the robot trajectory tracking control fuzzy PID controller through the following steps: generating an initial population, updating the population, updating the membership function and the fuzzy rules, calculating the target function value, judging whether to traverse N individual chromosomes, judging whether to terminate the genetic optimization algorithm, and generating the next generation of subgroups through genetic operations. The fuzzy PID controller after the mixed coding method genetic algorithm optimization reduces the probability of causing a decrease in control effect, reduces the tracking deviation, and reduces the impact on its own anti-interference ability. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A block diagram of a robot control system based on big data and artificial intelligence provided by the present invention;

[0032] Figure 2 A flow chart of a fuzzy PID controller for optimizing robot trajectory tracking control using a genetic algorithm based on hybrid coding provided by the present invention;

[0033] Figure 3 A flowchart of the genetic algorithm provided by the present invention for solving optimization problems. DETAILED DESCRIPTION

[0034] The present invention will be further described below in conjunction with the accompanying drawings and implementation modes.

[0035] Please refer to Figure 1 , Figure 2 as well as Figure 3 ,in, Figure 1 A block diagram of a robot control system based on big data and artificial intelligence provided by the present invention; Figure 2 A flow chart of a fuzzy PID controller for optimizing robot trajectory tracking control using a genetic algorithm based on hybrid coding provided by the present invention; Figure 3 A flowchart of the genetic algorithm provided by the present invention for solving optimization problems.

[0036] In the specific implementation process, Figure 1 , Figure 2 as well as Figure 3 As shown, it includes a data acquisition module, which is responsible for collecting various data from the robot's internal and external environment;

[0037] It should be noted that various data include sensor data, environmental parameters and user input data. Sensor data include vision, sound, radar, pressure, and light. Environmental parameters include temperature and humidity.

[0038] At the same time, the data acquisition module plays the role of "eyes" and "ears" in the robot control system. It obtains the robot's position, posture, and environmental key data in real time through various sensors and network equipment.

[0039] Data preprocessing module, used to clean, format and remove noise from the collected raw data;

[0040] It should be noted that data cleaning is the first step of data preprocessing, which aims to remove noise, duplication and missing unnecessary data in the data set to ensure the quality and accuracy of the data;

[0041] Remove duplicate data. Duplicate data may appear in the data set, which may affect the training and prediction of the model.

[0042] Remove outliers. Outliers are values ​​in a data set that are significantly different from other data. They may be caused by data recording errors, measurement errors, or other reasons.

[0043] Fill missing values. Missing values ​​may appear in data sets due to measurement errors, data entry errors or other reasons. In order to ensure the integrity and accuracy of the data, missing values ​​need to be filled. Common filling methods include mean filling, median filling, mode filling or model-based prediction filling.

[0044] Data transformation: Data transformation refers to converting raw data into a form more suitable for machine learning algorithms;

[0045] Feature scaling: Feature scaling refers to scaling down or up the feature values ​​so that they have the same order of magnitude, which can reduce the difference between feature values ​​and improve the performance of the model. Common feature scaling methods include minimum-maximum normalization (scaling the data to a range between 0 and 1) and Z-score normalization (scaling the data to a range with a mean of 0 and a standard deviation of 1).

[0046] Feature encoding is the process of converting categorical features into numerical features, because most machine learning algorithms require the input data to be numerical. This can be achieved through One-Hot Encoding, Label Encoding, or Target Encoding methods.

[0047] Feature selection: Feature selection is to select the most relevant features from all available features to reduce the number of features and improve the performance and interpretability of the model. Common feature selection methods include filtering methods (such as chi-square test, mutual information), wrapping methods (such as recursive feature elimination) and embedding methods (such as feature selection based on tree models);

[0048] A data storage module, used for storing the preprocessed data in a large-scale data storage system;

[0049] It should be noted that the functions of the data storage module are:

[0050] 1. Persistently store all kinds of data generated during the operation of the robot to ensure the security and reliability of the data;

[0051] 2. Provide data query, update, and deletion management functions to support fast access and efficient processing of stored data;

[0052] 3. Regularly back up stored data to prevent data loss or damage, and provide data recovery functions to ensure the integrity and availability of system data;

[0053] Data analysis and intelligent processing module, which uses big data processing technology and artificial intelligence algorithms to analyze and mine stored data;

[0054] It should be noted that the main components and functions of the data analysis and intelligent processing module are as follows:

[0055] 1. Data preprocessing, including data cleaning, data standardization and feature extraction;

[0056] 2. Data storage and management: Use relational databases or NoSQL databases (such as MongoDB) to store and manage data, and build data warehouses to integrate data from different sensors and external data sources;

[0057] 3. Data analysis: using statistical methods to conduct descriptive analysis of data;

[0058] 4. Machine learning and deep learning, supervised learning, using classification and regression algorithms (such as SVM, random forest, neural network) to train labeled data to achieve prediction and classification tasks, unsupervised learning, using clustering and dimensionality reduction (such as PCA, t-SNE) algorithms to analyze unlabeled data and discover structures and patterns in the data;

[0059] The control algorithm optimization module adopts a hybrid coding method to encode the design parameters and characteristic parameters of the fuzzy PID controller, and uses a genetic algorithm to optimize the membership function and fuzzy rules of the fuzzy PID controller. At the same time, the individuals are genetically operated according to the fitness function corresponding to the selected objective function, and when running the genetic algorithm, each design parameter is directly represented by the individual chromosome gene position. Direct coding method;

[0060] It should be noted that when using genetic algorithms, a direct encoding method is used to directly represent each design parameter with individual chromosome gene positions;

[0061] At the same time, the optimization of robot trajectory tracking control fuzzy PID controller through hybrid coding-based genetic algorithm includes the following steps:

[0062] S1. Generate the initial population. According to the range of the characteristic parameters and design parameters of the fuzzy controller, adopt the mixed encoding method of characteristic parameters and design parameters, and randomly generate N individual chromosomes [a1, a2, a3, ..., a 63 ];

[0063] It should be noted that when generating the initial population, it is first necessary to clarify which data in the fuzzy PID controller needs to be optimized. The characteristic parameters of the fuzzy PID controller usually include the division of fuzzy sets (such as fuzzy subsets and membership functions of input and output variables) and fuzzy rule bases (i.e., IF-THEN rules), while the design parameters mainly refer to the three parameters of proportion (P), integral (I), and differential (D) in the PID controller. When determining which data in the fuzzy PID controller needs to be optimized, the parameter value range is set, and a reasonable value range is set for each parameter that needs to be optimized. Then, a mixed coding method is set. After determining the coding method and the parameter value range, N individuals can be randomly generated as the initial population.

[0064] S2, update the population, update the current population to the parent population;

[0065] It should be noted that genetic information is exchanged between selected parent individuals to produce new offspring. For mixed coding, you may need to design different crossover operations for different types of coding, and change certain parts of the individual coding with a certain probability to increase the diversity of the population. Similarly, the mutation operation also needs to be adjusted according to the coding type;

[0066] S3, update the membership function and fuzzy rules, read the gene codes of the jth chromosome in the current population, and calculate the gene codes according to the codes a1-a 10 Generate new input variables e, ec and output variable ΔK P , ΔK I , ΔK D Membership function, update fuzzy rules;

[0067] It should be noted that when using the genetic algorithm, a direct encoding method is used to directly represent each design parameter with the gene bit of the individual chromosome. The gene bit generates the central vertex values ​​of the membership function of the new input variables e, ec and the output variables,,, and each variable needs to specify 3 central vertex values. Therefore, 5 variables require 15 gene bits to represent the design parameters of the membership function of the fuzzy PID controller. The fuzzy PID controller has a total of three rule tables, and each rule table requires 49 fuzzy rule results to be designed;

[0068] S4, calculate the objective function value, run the fuzzy PID controller to complete the trajectory tracking control simulation, and calculate the deviation e(t) between the simulated trajectory of the robot and the expected trajectory at time t, and calculate the objective function value J;

[0069] It should be noted that the objective function of the controller designed by selecting the ITAE criterion is as follows:

[0070]

[0071] Where, e(t) is the deviation between the simulated trajectory of the robot at time t and the expected trajectory, T is the total time for the motion fuzzy PID controller to complete a trajectory tracking control simulation, and the fitness function of the genetic algorithm is F = 1 / J;

[0072] S5, determine whether to traverse N individual chromosomes. If the number of individual chromosomes j does not satisfy j>N, then j+1 and enter S3; if j>N, then enter S6;

[0073] S6, judging whether to terminate the genetic optimization algorithm, judging whether the change of the objective function value is less than the set threshold or whether the genetic generation reaches the set value, if satisfied, then terminate the genetic optimization algorithm, if not satisfied, then enter S7;

[0074] S7, Genetic operation to generate the next generation of sub-populations: According to the fitness function value converted from the objective function value of each individual in the current population, a new population can be obtained by performing basic genetic algorithm operations such as selection, replication, crossover and mutation on all individuals in the population, and then enter S2;

[0075] At the same time, the basic operations of genetic algorithms are divided into the following categories:

[0076] 1) Selection: The selection operation simulates the phenomenon of survival of the fittest in nature. Individuals with high fitness values ​​are more likely to be selected to produce one or more offspring, while individuals with low fitness values ​​may also be selected as candidates;

[0077] It should be noted that the selection methods include fitness ratio method, ranking method and roulette method;

[0078] 2) Replication: The replication operation simulates the phenomenon of survival of the fittest, that is, several individuals with the highest fitness in the parent group have a probability of being directly copied into the child group, which ensures the improvement of the optimal fitness value of each generation of the population;

[0079] It should be noted that several individuals with the highest fitness values ​​are called elites, the number of elites is 1 to 4 or 0.05 of the population, and the replication probability is 0.8;

[0080] 3) Crossover: The crossover operation simulates the reproduction process of organisms in nature. It generates the next generation of individual chromosomes by exchanging and recombining the genetic information of the two individual chromosomes of the parent generation.

[0081] It should be noted that the commonly used crossover operations include single-point crossover, two-point crossover and arithmetic crossover. For single-point crossover and two-point crossover, one or more points on the parent chromosomes are randomly selected to exchange parts on one side of the point to obtain a new chromosome number string. For continuous real-valued genes, the formula for arithmetic crossover can be used as follows:

[0082] X' ij =r j x 1j +(1-r)x 2j

[0083] In the formula, r j is a random value uniformly distributed in the interval (0, 1), x 1j and x 2j are the j-th genes in the chromosomes of the parents respectively;

[0084] 4) Mutation: The mutation operation simulates the gene mutation caused by accidental factors in the process of biological evolution. Each gene position on the individual chromosome has a very small probability of randomly mutating to other values.

[0085] It should be noted that the value of the gene bit on the individual chromosome of the genetic algorithm can be a discrete value or a continuous real number. When the value of the gene bit is binary, the mutation operation is to invert the random gene bit. When the value of the gene bit has more than two discrete values, the mutation operation is to select one from the optional discrete values. When the value of the gene bit is a continuous real value, the mutation operation is to add a random value that obeys the normal distribution with zero mean and small variance to the current value.

[0086] In addition, the components of genetic algorithm mainly include:

[0087] (1) Chromosome encoding scheme: Each chromosome of GA is a bit string. The number of bits in the bit string is determined by the number of decision variables in the optimization problem and the encoding method.

[0088] It should be noted that there are two types of chromosome encoding: binary encoding and real number encoding. For continuous real-valued variables, real number encoding is used, and for discrete variables, binary encoding is used. The process of mapping chromosome bit strings to individual phenotypes is called decoding. The decoding method determines the transformation rules between chromosomes and solutions to optimization problems.

[0089] (2) Individual objective function and fitness function. GA uses the objective function as search information and the fitness function value as the internal driving force for optimization.

[0090] It should be noted that the correct selection of the objective function of the optimization problem and the appropriate selection of the transformation rule between the objective function and the fitness function will greatly affect the optimization results of the GA algorithm, so it is necessary to determine the calculation method of the objective function and fitness function of the optimization problem;

[0091] (3) Selection of basic operations of genetic algorithm. The basic operations of GA include selection, replication, crossover and mutation. It is necessary to reasonably select the methods of selection, crossover and mutation. Different methods of different genetic operations will produce different optimization effects.

[0092] It should be noted that for continuous real-valued variables, the selection operation is the roulette method, the crossover operation uses arithmetic crossover, and the mutation operation uses the continuous real-valued gene method;

[0093] (4) Parameters required for running the genetic algorithm. After determining the methods of different genetic operations, it is still necessary to set the necessary parameters in the method and other parameters required for running the GA;

[0094] It should be noted that the necessary parameters for GA operation include population size, termination generation of genetic algorithm or threshold of fitness value change;

[0095] The specific principle of the fuzzy PID controller is: first, the deviation e and deviation change ec of the controlled variable are used as the two input variables of the fuzzy controller. After fuzzification, the membership degree of e and ec belonging to different fuzzy sets is obtained. Then, fuzzy reasoning is performed based on fuzzy rules to obtain the output variable ΔK. P , ΔK I , ΔK D The fuzzy result of the output variable is obtained by defuzzification method;

[0096] The design of a genetic algorithm to solve an optimization problem consists of the following steps:

[0097] S10, determining the parameters to be optimized and the optimization constraints such as the upper and lower limits of the parameters according to the optimization problem;

[0098] S20, designing a chromosome encoding representation method to achieve the mutual conversion relationship between the parameters to be optimized and the chromosome gene positions;

[0099] S30, determining a calculation method of an objective function and a fitness function of a genetic algorithm according to the target to be optimized;

[0100] S40, selecting basic operations of the genetic algorithm, and determining specific implementation methods of selection, replication, crossover and mutation;

[0101] S50, setting necessary parameters in the specific implementation method of the genetic operation in S40, such as crossover probability and mutation probability, and setting other genetic operation parameters such as population size and optimization termination condition;

[0102] At the same time, the design parameters of the fuzzy PID controller specifically refer to the central vertex position of the membership function of the input and output variables and the output rule then result obtained by each fuzzy rule if condition. For the characteristic parameters of the membership function, according to the characteristics of the membership function design, its expression is as follows:

[0103]

[0104] In the formula, X mi is the density between the central vertex position of the triangle membership function and zero, k is the number of fuzzy sets of the membership function, where k = 7, p m is the exponent of the power function, p m The value will affect the density distribution of each central value of the membership function, and based on this expression, the scaling factor G can be further used to describe the density function of the vertices of the triangle membership function as follows:

[0105]

[0106] The p m and G are defined as the characteristic parameters of the membership function, then the design parameters of the membership function X i Can be obtained by p m and G generation, that is, the three trigonometric function vertex positions required for designing the membership function can be generated by two characteristic parameters;

[0107] Moreover, ΔK in the fuzzy rule P and ΔK D The output rules of can be divided into several regions with the same rule results. The number of regions depends on the number k of fuzzy sets of the output variable membership function and ΔK P The fuzzy rule of ΔK is used as an example to illustrate the method of using phase plane to divide regions. PThe corresponding fuzzy set values ​​of input variables e and ec are {NL, NM, NS, ZE, PS, PM, PL} to establish the phase plane XOY. The intersection points of different fuzzy set values ​​of X and Y axes are grid points. It can be seen that the grid points represent the if part of the fuzzy rule. The distance of the grid points from the X and Y axes represents the central vertex position mi X value of the triangular membership function corresponding to the different fuzzy set values ​​of e and ec. The coordinate calculation formula of the segmentation point is as follows:

[0108]

[0109] In the formula, (X si , Y si ) is the coordinate of the split point, L is the range parameter used to limit the split point to fall within the phase plane, sign is the sign function, θ s is the angle of the line connecting the split points rotating counterclockwise around the origin along the positive X-axis, P s is the power function exponent. Finally, a set of parallel straight lines are generated through the perpendicular bisector between two adjacent split points. These parallel lines divide the phase plane into regions. At the same time, the fuzzy output rule results corresponding to the grid points in the same region are specified to be the same, that is, the fuzzy rule results are the same. According to the PID controller parameter K P , K I , K D The general principle of adjustment is that when e is large, ΔK P Should take a smaller value, when e is smaller, ΔK P Take a smaller value and select ΔK appropriately D When ec is large, ΔK D Take the smaller value, otherwise ΔK D Taking a larger value, ΔK P The results of the regular division of the phase plane area are designated as NL, NM, NS, ZE, PS, PM, PL from left to right;

[0110] The control decision module is used to formulate the robot's control strategy and instructions based on the output results of the data analysis and intelligent processing module;

[0111] It should be noted that it is responsible for data analysis and decision-making through algorithms and models based on the environmental information obtained by the perception module to determine the plan and strategy of the robot's behavior;

[0112] The execution module is responsible for converting the instructions formulated by the control decision module into the actual actions of the robot;

[0113] It should be noted that the execution module is usually composed of an actuator, a transmission mechanism, a sensor and a feedback system. The actuator is the physical basis of the execution module. It converts electrical energy, hydraulic energy or air pressure energy into mechanical energy to drive the robot joint movement or perform end operations.

[0114] The transmission mechanism is responsible for transmitting the power of the actuator to various parts of the robot to ensure the accurate execution of the action. It may include gears, chains, belts, mechanical transmission elements, as well as reducers and coupling auxiliary devices;

[0115] Sensors are used to monitor the robot's motion state and external environment changes in real time, such as position, speed, acceleration, force, and temperature. The feedback system adjusts the output of the actuator based on the sensor data to ensure that the robot moves according to the predetermined trajectory and parameters.

[0116] The circuits and controls involved in the present invention are all prior art and will not be described in detail here.

[0117] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any effective structural or effective process changes made using the contents of the present invention's specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A robot control system based on big data and artificial intelligence, characterized in that: It includes a data acquisition module, which is responsible for collecting various data from the robot's internal and external environment; A data preprocessing module is used to perform preprocessing operations on the collected raw data; A data storage module, used for storing the preprocessed data in a large-scale data storage system; Data analysis and intelligent processing module, which uses big data processing technology and artificial intelligence algorithms to analyze and mine stored data; The control algorithm optimization module adopts a hybrid coding method to encode the design parameters and characteristic parameters of the fuzzy PID controller, and uses a genetic algorithm to optimize the membership function and fuzzy rules of the fuzzy PID controller. At the same time, the individuals are genetically operated according to the fitness function corresponding to the selected objective function, and when running the genetic algorithm, each design parameter is directly represented by the individual chromosome gene position. Direct coding method; The control decision module is used to formulate the robot's control strategy and instructions based on the output results of the data analysis and intelligent processing module; The execution module is responsible for converting the instructions given by the control decision module into actual actions of the robot.

2. The robot control system based on big data and artificial intelligence according to claim 1 is characterized in that: Optimizing the robot trajectory tracking control fuzzy PID controller through hybrid coding based genetic algorithm includes the following steps: S1, generate the initial population, according to the characteristic parameters of the fuzzy controller and the value range of the design parameters, adopt the characteristic parameter and design parameter mixed encoding method, and randomly generate N individual chromosomes; S2, update the population, update the current population to the parent population; S3, update the membership function and fuzzy rules, read the gene codes of the jth chromosome in the current population, and 10 Generate new input variables e, ec and output variable ΔK P , ΔK I , ΔK D Membership function, update fuzzy rules; S4, calculate the objective function value, run the fuzzy PID controller to complete the trajectory tracking control simulation, and calculate the deviation e(t) between the simulated trajectory of the robot and the expected trajectory at time t, and calculate the objective function value J; S5, determine whether to traverse N individual chromosomes. If the number of individual chromosomes j does not satisfy j>N, then j+1 and enter S3; if j>N, then enter S6; S6, judging whether to terminate the genetic optimization algorithm, judging whether the change of the objective function value is less than the set threshold or whether the genetic generation reaches the set value, if satisfied, then terminate the genetic optimization algorithm, if not satisfied, then enter S7; S7, Genetic operations generate the next generation of sub-populations: Based on the fitness function value converted from the objective function value of each individual in the current population, a new population can be obtained by performing basic genetic algorithm operations such as selection, replication, crossover and mutation on all individuals in the population, and then enter S2.

3. The robot control system based on big data and artificial intelligence according to claim 2 is characterized in that: The specific control process of the fuzzy PID controller is as follows: first, the deviation e and deviation change ec of the controlled variable are taken as the two input variables of the fuzzy controller. After fuzzification, the membership degree of e and ec belonging to different fuzzy sets is obtained. Then, fuzzy reasoning is performed based on fuzzy rules to obtain the output variable ΔK. P , ΔK I , ΔK D The fuzzy result is defuzzified and the precise result of the output variable is obtained.

4. The robot control system based on big data and artificial intelligence according to claim 3 is characterized in that: The design of a genetic algorithm to solve an optimization problem consists of the following steps: S10, determining the parameters to be optimized and the optimization constraints such as the upper and lower limits of the parameters according to the optimization problem; S20, designing a chromosome encoding representation method to achieve the mutual conversion relationship between the parameters to be optimized and the chromosome gene positions; S30, determining a calculation method of an objective function and a fitness function of a genetic algorithm according to the target to be optimized; S40, selecting basic operations of the genetic algorithm, and determining specific implementation methods of selection, replication, crossover and mutation; S50, setting necessary parameters in the specific implementation method of the genetic operation in S40, such as crossover probability and mutation probability, and setting other genetic operation parameters such as population size and optimization termination condition.

5. The robot control system based on big data and artificial intelligence according to claim 4 is characterized in that: The design parameters of the fuzzy PID controller specifically refer to the central vertex position of the membership function of the input and output variables and the output rule then result obtained by each fuzzy rule if condition.

6. The robot control system based on big data and artificial intelligence according to claim 5 is characterized in that: The various data include sensor data, environmental parameters and user input data.