Multi-stage control method for micro servo electric cylinder

Abstract

The invention relates to the technical field of multi-stage control, and particularly discloses a multi-stage control method for a micro servo electric cylinder, which comprises the following steps of: setting a monitoring time window, and acquiring temperature and load data of the electric cylinder in real time; calculating the normalized deviation of each index and a reference value and carrying out weighted fusion to obtain a real-time disturbance index; analyzing the thermal management situation and the mechanical layout stability of the electric cylinder, and evaluating and generating an ontology state index based on a preset threshold value; constructing a decision rule base realized in a two-dimensional look-up table form, taking the ontology state index as input, mapping and outputting a preset combination of a parameter adjustment instruction and a performance constraint instruction, and adjusting the operating parameters of the electric cylinder according to the combination to realize closed-loop control; the problem that the control precision is reduced in a multi-physical-quantity coupling environment is effectively solved, and prospective state evaluation and self-adaptive adjustment are carried out.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of multi-stage control, and particularly relates to a multi-stage control method for a micro servo cylinder. BACKGROUND

[0002] As a precise linear motion actuator, the micro servo cylinder has been widely used in industrial automation, robots, precision machining and other fields. Its control performance directly affects the positioning accuracy, response speed and running stability of the whole system.

[0003] The traditional micro servo cylinder control method mainly relies on the PID control algorithm with fixed parameters, which can meet the basic control requirements to a certain extent, but has obvious limitations in dealing with complex and variable working environments. In actual operation, the micro servo cylinder is affected by temperature fluctuations and load fluctuations: temperature changes will cause motor parameter drift and mechanical structure thermal deformation; load fluctuations will change the system dynamics and cause control deviation. The existing technology often uses single temperature compensation or load observation method, lacks a comprehensive evaluation mechanism for the coupling influence of multiple physical quantities, resulting in a lack of systematicness in control decision-making. At the same time, the traditional scheme focuses on real-time adjustment in operation, and ignores the innate constraints of the cylinder body state on control performance. This passive control mode is difficult to maintain optimal control performance in fluctuating environments, resulting in decreased control accuracy, deteriorated response characteristics, and even causing system oscillation or overload failure.

[0004] Therefore, there is an urgent need for a new control method that can comprehensively evaluate internal and external factors, implement forward-looking adjustment, and combine real-time adaptive control. SUMMARY

[0005] The purpose of the present application is to provide a multi-stage control method for a micro servo cylinder, which solves the following technical problems.

[0006] The purpose of the present application can be achieved by the following technical solutions: A multi-stage control method for a micro servo cylinder, comprising the following steps: Step S1: setting a monitoring time window, in which real-time acquisition of a plurality of index data of the micro servo cylinder is performed, the index data including temperature data and load data; presetting reference values of each index data, normalizing the deviation of the index data according to the reference values of the index data, and weighting and fusing each index data after normalization to obtain a real-time disturbance index; Step S2: generating a numerical regression line of the index data in the monitoring time window, obtaining the change rate and change span of the index data based on the numerical regression line, and obtaining a stability index of the index data; based on the stability index of each index data, obtaining an intrinsic state index of the micro servo cylinder, denoted as the current intrinsic state index; Step S3: a plurality of instruction combinations are preset, the instruction combinations include parameter adjustment instructions and performance constraint instructions; a decision rule base is established based on a two-dimensional query table, with the ontology state index as input and the instruction combination as output; the current ontology state index is input into the decision rule base to obtain a decision instruction, and the parameter of the micro servo cylinder is adjusted according to the decision instruction.

[0007] As a further scheme of the present application: the reference value of the index data is a rated working parameter of the index corresponding to the index data; The index corresponding to the index data can be understood as the temperature of the temperature data, and the reference value of the temperature data is the rated working temperature, so the reference value of the load data is the rated output thrust.

[0008] As a further scheme of the present application: the real-time disturbance index , wherein w i represents the preset weight of the i-th index data, Q i (t) represents the real-time value of the i-th index data, Q i ´ represents the reference value of the i-th index data, Q i represents the index range of the i-th index data, N is the total number of index data, i ∈ [1, N] and i is a positive integer; The index range of the index data is set based on the rated working range of the index corresponding to the index data, and the index range is the difference between the maximum value and the minimum value of the rated working range.

[0009] As a further scheme of the present application: the weight of the index data is set based on the entropy weight method, and the process includes: For any index data, historical sample data is obtained, the historical sample data includes historical index data of a plurality of micro servo cylinder samples, and the historical index data is a plurality of index values; a m×n-dimensional historical data matrix is constructed based on the historical sample data, wherein m represents the number of micro servo cylinder samples, and n represents the number of index values of any micro servo cylinder sample; the historical data matrix is standardized to obtain a standardized matrix Y, in which the j-th index value of the k-th micro servo cylinder sample is standardized to obtain an element y kj =(x kj -Minx j ) / (Maxx j -Minx j ), wherein x kj represents the j-th index value of the k-th micro servo cylinder sample, Minx j represents the minimum value of the index value, and Maxx jThe maximum value of the index value is represented; the proportion of the jth index value of the kth micro servo cylinder sample is calculated The entropy value of the jth index value is calculated The difference value d of the jth index value is calculated j =1-e j The weight of the index data is obtained .

[0010] As a further scheme of the present application: the generation process of the numerical regression line comprises: A coordinate system is established with a timestamp as the abscissa and the numerical value of the index data as the ordinate; the numerical value of the real-time collected index data and the corresponding timestamp when the numerical value is collected are converted into coordinate points at the corresponding positions in the coordinate system, and a regression straight line of each coordinate point is obtained based on the least square method, which is denoted as the numerical regression line.

[0011] As a further scheme of the present application: the change rate of the index data is the slope of the numerical regression line; The average value of all numerical values in the index data is obtained, the absolute value of the difference between any numerical value and the average value is obtained, and the average value of all absolute values is obtained, which is denoted as the change span of the index data.

[0012] As a further scheme of the present application: the stability index S of the index data is S=1 / (1+E1|K|+E2σ), wherein K is the change rate of the index data, σ is the change span of the index data, and E1 and E2 are both preset adjustment coefficients; The body state index of the micro servo cylinder , wherein S i represents the stability index of the ith index data.

[0013] As a further scheme of the present application: the setting process of the instruction combination comprises: A plurality of running conditions are set, each running condition corresponding to an instruction combination, the running conditions comprising a smooth running condition, a load mutation condition, a temperature abnormality condition and a composite abnormality condition; the parameter adjustment instruction and the performance constraint instruction in the running condition are based on artificial configuration; The parameter adjustment instruction comprises the adjustment amount of the gain parameter of the PID controller, and the adjustment amount of the gain parameter comprises the proportional coefficient adjustment amount, the integral coefficient adjustment amount and the differential coefficient adjustment amount; the performance constraint instruction comprises the maximum speed limit value, the maximum thrust limit value and the acceleration limit value.

[0014] The present application has the following beneficial effects: The application quantifies the influence of external temperature and load fluctuation by real-time disturbance index, accurately evaluates the stability of internal thermal management and mechanical layout of the electric cylinder by body state index, and establishes a complete technical chain from data acquisition, state evaluation to intelligent decision-making; the decision rule base based on two-dimensional query table is adopted to realize the collaborative adjustment of control parameters and performance constraints, which not only ensures the rapid response ability of the control system, but also ensures the operation safety; compared with the traditional fixed parameter PID control, the application effectively solves the problem of control precision decline in the environment of multiple physical quantity coupling, and through the forward-looking state evaluation and adaptive adjustment, the control quality and reliability of the system under fluctuating working conditions are significantly improved, and the modular design idea facilitates engineering implementation and maintenance. BRIEF DESCRIPTION OF DRAWINGS

[0015] The application will be further described below with reference to the drawings.

[0016] Figure 1 is a step schematic diagram of a multi-stage control method for a micro servo cylinder. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0018] Please refer to Figure 1 The application is a multi-stage control method for a micro servo cylinder, comprising the following steps: Step S1: setting a monitoring time window, in the monitoring time window, a plurality of index data of the micro servo cylinder are collected in real time, the index data include temperature data and load data; reference values of the index data are preset, the index data are normalized deviation according to the reference values of the index data, and each index data after normalization deviation is weighted and fused to obtain a real-time disturbance index; As a preferred embodiment of the application, the temperature data is obtained based on a temperature sensing device, and the load data is obtained based on a mechanical sensing device; As a preferred embodiment of the application, the reference value of the index data is set as the rated working parameter of the index corresponding to the index data; The index corresponding to the index data can be understood as the temperature of the temperature data, so the reference value of the temperature data is the rated working temperature, and the reference value of the load data is the rated output thrust; As a preferred embodiment of the application, the real-time disturbance index , where w i Q represents the preset weight of the i-th indicator data. i (t) represents the real-time value of the i-th indicator data, Q i ´ represents the reference value of the i-th indicator data, Q i Let N represent the range of the i-th indicator data, N be the total number of indicator data, i ∈ [1, N] and i is a positive integer; The range of the indicator data is set based on the rated working range of the indicator corresponding to the indicator data, and the range of the indicator is the difference between the maximum and minimum values ​​of the rated working range. Specifically This step is the normalized deviation calculation, which eliminates the dimensional differences between different indicators, enabling direct comparison and fusion of different physical quantities such as temperature and load; the real-time disturbance index M(t) is a multi-indicator weighted fusion result, which obtains a dimensionless comprehensive evaluation value by comprehensively processing multiple key operating parameters; a positive value indicates that the overall value is too high, and a negative value indicates that the overall value is too low. The weights of the indicator data are set based on the entropy weight method, and the process includes: For any given index data, historical sample data is obtained. This historical sample data includes historical index data from several micro servo cylinder samples, and each historical index data consists of several index values. An m×n dimensional historical data matrix is ​​constructed based on this historical sample data, where m represents the number of micro servo cylinder samples and n represents the number of index values ​​for any given micro servo cylinder sample. The historical data matrix is ​​then standardized to obtain a standardized matrix Y. In this standardized matrix, the element y is the standardized value of the j-th index of the k-th micro servo cylinder sample. kj =(x kj -Minx j ) / (Maxx j -Minx j ), where x kj Minx represents the value of the j-th index of the k-th miniature servo electric cylinder sample. j Maxx represents the minimum value of the indicator. j This represents the maximum value of the index; therefore, the proportion of the j-th index value of the k-th micro servo electric cylinder sample is calculated. The entropy value of the j-th index is calculated. And calculate the difference value d of the j-th indicator. j =1-e j Then the weights of the indicator data are obtained. ; Step S2: Generate a numerical regression line for the indicator data within the monitoring time window. Based on the numerical regression line, obtain the rate of change and span of change of the indicator data, and obtain the stability index of the indicator data. Based on the stability index of each indicator data, obtain the body state index of the micro servo electric cylinder, which is denoted as the current body state index. In a preferred embodiment of the present invention, the process of generating the numerical regression line includes: A coordinate system is established with timestamps as the horizontal axis and the values ​​of indicator data as the vertical axis. The values ​​of the indicator data collected in real time and the timestamps corresponding to the collection time are converted into coordinate points at corresponding positions on the coordinate system. The regression lines of each coordinate point are obtained based on the least squares method and are denoted as numerical regression lines. In a preferred embodiment of the present invention, the rate of change of the index data is the slope of the numerical regression line; In a preferred embodiment of the present invention, the process of obtaining the range of change of the indicator data includes: Obtain the average value of all values ​​within the indicator data. For any value, obtain the absolute value of the difference between the value and the average value, and obtain the average value of all absolute values. This is recorded as the range of change in the indicator data. In a preferred embodiment of the present invention, the stability index S of the indicator data is S=1 / (1+E1|K|+E2σ), where K is the rate of change of the indicator data, σ is the range of change of the indicator data, and E1 and E2 are preset adjustment coefficients. When K or σ increases, the value of S decreases, which is consistent with the physical intuition that the greater the fluctuation, the worse the stability. The value of S is in the range of (0,1], which facilitates the comparison and integration of different indicators. The adjustment coefficient is used to balance the influence of the rate of change and the span of change on the stability index. The adjustment coefficient is set based on historical data optimization. Historical data samples of the micro servo electric cylinder under stable operation and abnormal conditions are collected. The coefficient value that makes the stability index best distinguish between different states is derived through regression analysis or optimization algorithm, and E1+E2=1. It is worth noting that the absolute value of the rate of change is used to avoid the cancellation of positive and negative values, and the range of change reflects the degree of data dispersion. In a preferred embodiment of the present invention, the body state index of the micro servo electric cylinder S i This represents the stability index of the i-th indicator data; It is understandable that the body state index uses a geometric average algorithm to integrate the health of multiple independent indicators into a single, sensitive and reliable comprehensive score. The result is a value between 0 and 1, and the closer it is to 1, the healthier and more stable the electric cylinder body state is. Step S3: Preset several instruction combinations, including parameter adjustment instructions and performance constraint instructions; establish a decision rule base based on a two-dimensional lookup table, with the body state index as input and the instruction combination as output; input the current body state index into the decision rule base to obtain decision instructions, and adjust the parameters of the micro servo electric cylinder according to the decision instructions; In a preferred embodiment of the present invention, the process of setting the instruction combination includes: Several operating conditions are set, and each operating condition corresponds to a combination of instructions. The operating conditions include stable operation, sudden load change, abnormal temperature, and combined abnormal conditions. The parameter adjustment instructions and performance constraint instructions within the operating conditions are based on manual configuration. The parameter adjustment instructions include the adjustment amount of the gain parameter of the PID controller, which includes the adjustment amount of the proportional coefficient, integral coefficient, and derivative coefficient; the performance constraint instructions include the maximum speed limit, the maximum thrust limit, and the acceleration limit. In a preferred embodiment of the present invention, the process of establishing the decision rule base includes: The continuous body state index and real-time disturbance index are discretized into several levels. The body state index and the real-time disturbance index of each level are randomly combined in pairs to obtain several state combinations. Each state combination maps to an operating condition. It is worth noting that, based on historical data analysis, the state combination with the best control effect under the operating conditions is selected as the input corresponding to the operating conditions; In a preferred embodiment of the present invention, the process of adjusting the parameters of the micro servo electric cylinder includes: The parameter adjustment instructions in the decision instructions are mapped to the adjustment amounts of various gain parameters of the PID controller, and the operating parameters of the PID controller are updated based on each adjustment amount; and the performance constraint instructions in the decision instructions are converted into real-time limits on the micro servo electric cylinder. The conversion process includes setting the maximum allowable speed of the micro servo electric cylinder to the maximum speed limit value in the performance constraint instructions, and setting the maximum output thrust of the micro servo electric cylinder to the maximum thrust limit value in the performance constraint instructions. Specifically, the proportional coefficient adjustment is denoted as Kp, the integral coefficient adjustment as Ki, and the derivative coefficient adjustment as Kd; and the initial value of the proportional coefficient adjustment at the factory setting, Kp0, and the initial value of the integral coefficient adjustment at the factory setting, Ki, are obtained. 0， The initial value of the differential coefficient adjustment amount when it is set at the factory; At this point, four discrete levels are set for the real-time disturbance index M(t): stable, mild, moderate, and severe; four discrete levels are set for the body state index H: excellent, satisfactory, warning, and abnormal; then, parameter adjustment configurations corresponding to 16 state combinations are established, for example: When (H,M(t))=(good, stable), Kp=+0.1Kp0, Ki=+0.05Ki0, Kd=-0.02Kd0, at this time, the proportional and integral actions are slightly increased, and the differential term is finely adjusted; When (H,M(t))=(qualified, moderate), Kp=+0.3Kp0,Ki=+0.15Ki0,Kd=+0.1Kd0, at this time the gains of each item are significantly enhanced to cope with moderate disturbances; After obtaining the adjustment amount corresponding to the decision command, update the PID controller parameters according to the following formula: The adjustment amount of the updated scaling factor is Kp new =Kp now ×(1+Kp), where Kp now The current proportional coefficient adjustment is updated to Ki. new =Ki now ×(1+Ki), where Ki now The current integral coefficient adjustment is used to update the differential coefficient adjustment to Kd. new =Kd now ×(1+Kd), where Kd now This is the current adjustment amount for the differential coefficients.

[0019] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the scope of the invention.

Claims

1. A multi-stage control method for a micro-servo cylinder, characterized by, The method comprises the following steps: Step S1: setting a monitoring time window, collecting several index data of the micro servo cylinder in real time within the monitoring time window, the index data including temperature data and load data; presetting reference values of each index data, normalizing the deviation of the index data according to the reference values of the index data, and weighting and fusing each index data after normalization to obtain a real-time disturbance index; Step S2: generating a numerical regression line of the index data within the monitoring time window, obtaining the change rate and change span of the index data based on the numerical regression line, and obtaining a stability index of the index data; obtaining a body state index of the micro servo cylinder based on the stability index of each index data, denoted as a current body state index; Step S3: presetting several instruction combinations, the instruction combination including a parameter adjustment instruction and a performance constraint instruction; establishing a decision rule library based on a two-dimensional query table, taking the body state index as the input and the instruction combination as the output; inputting the current body state index into the decision rule library to obtain a decision instruction, and adjusting the parameters of the micro servo cylinder according to the decision instruction.

2. A multi-stage control method for a micro servo cylinder according to claim 1, characterized in that, In step S1, the reference value of the index data is the rated working parameter of the index corresponding to the index data; The index corresponding to the index data can be understood as the temperature of the temperature data, so the reference value of the temperature data is the rated working temperature, and the reference value of the load data is the rated output thrust.

3. A multi-stage control method for a micro servo cylinder according to claim 1, characterized in that, In step S1, the real-time disturbance index where w i represents the preset weight of the i-th index data, Q i (t) represents the real-time value of the i-th index data, Q i represents the reference value of the i-th index data, Q i represents the index range of the i-th index data, N is the total number of index data, i ∈ [1, N] and i is a positive integer; The index range of the index data is set based on the rated working range of the index corresponding to the index data, and the index range is the difference between the maximum and minimum values of the rated working range.

4. A multi-stage control method for a micro servo cylinder according to claim 3, wherein, In step S1, the weight of the index data is set based on the entropy weight method, and the process includes: For any given index data, historical sample data is obtained. This historical sample data includes historical index data from several micro servo cylinder samples, and each historical index data consists of several index values. An m×n dimensional historical data matrix is ​​constructed based on this historical sample data, where m represents the number of micro servo cylinder samples and n represents the number of index values ​​for any given micro servo cylinder sample. The historical data matrix is ​​then standardized to obtain a standardized matrix Y. In this standardized matrix, the element y is the standardized value of the j-th index of the k-th micro servo cylinder sample. kj =(x kj -Minx j ) / (Maxx j -Minx j ), where x kj Minx represents the value of the j-th index of the k-th miniature servo electric cylinder sample. j Maxx represents the minimum value of the indicator. j This represents the maximum value of the index; therefore, the proportion of the j-th index value of the k-th micro servo electric cylinder sample is calculated. The entropy value of the j-th index is calculated. And calculate the difference value d of the j-th indicator. j =1-e j Then the weights of the indicator data are obtained. .

5. A multi-stage control method for a micro-servo cylinder according to claim 1, wherein, In step S2, the generation process of the numerical regression line includes: establishing a coordinate system with a timestamp as the horizontal coordinate and the numerical value of the index data as the vertical coordinate; converting the numerical value of the real-time collected index data and the corresponding timestamp into a coordinate point at the corresponding position in the coordinate system, and obtaining a regression straight line of each coordinate point based on the least square method, denoted as a numerical regression line.

6. A multi-stage control method for a micro servo cylinder according to claim 1, wherein, In step S2, the change rate of the index data is the slope of the numerical regression line; obtaining the average value of all numerical values in the index data, obtaining the absolute value of the difference between the numerical value and the average value for any numerical value, and obtaining the average value of all absolute values, denoted as the change span of the index data.

7. A multi-stage control method for a micro-servo cylinder according to claim 1, wherein, In step S2, the stability index S of the index data is S=1 / (1+E1|K|+E2σ), where K is the change rate of the index data, σ is the change span of the index data, and E1 and E2 are both preset adjustment coefficients; Body condition index for micro servo cylinders where S i denotes the stability index of the i-th index data.

8. A multi-stage control method for a micro servo cylinder according to claim 1, wherein, In step S3, the setting process of the instruction combination includes: setting several operating conditions, each operating condition corresponding to an instruction combination, the operating conditions including a smooth operating condition, a load mutation operating condition, a temperature abnormality operating condition and a composite abnormality operating condition; the parameter adjustment instruction and the performance constraint instruction in the operating condition are configured artificially; The parameter adjustment instruction comprises an adjustment amount of a gain parameter of a PID controller, and the adjustment amount of the gain parameter comprises a proportional coefficient adjustment amount, an integral coefficient adjustment amount and a differential coefficient adjustment amount; and the performance constraint instruction comprises a maximum speed limit value, a maximum thrust limit value and an acceleration limit value.

Images