A guidance optimization method and device for a telescopic mechanism

By training multi-level stable identification model and optimizing the guidance distance, the problems of deviation and instability of the telescopic mechanism during the guidance process are solved, and higher guidance stability and control accuracy are achieved.

CN118780180BActive Publication Date: 2024-12-06JIANGSU GELANWEI INTELLIGENT EQUIP CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411258949.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-10
Publication Date
2024-12-06
Estimated Expiration
2044-09-10

AI Technical Summary

Technical Problem

The existing telescopic mechanisms rely on fixed structural parameters and control algorithms during the guidance process, making it difficult to adapt to different working conditions, resulting in bias and instability during the guidance process.

Method used

By determining the multi-stage telescopic parts of the telescopic mechanism, collecting structural parameters and sensing sample data, training a multi-stage stable identification model, obtaining the target guidance distance, and using the model to find optimization until the stability requirements are met, and the optimized guidance is output. Distance to control the telescopic mechanism.

Benefits of technology

The guidance stability and control accuracy of the telescopic mechanism are improved, and can better adapt to various working environments and task requirements.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118780180B_ABST
    Figure CN118780180B_ABST
Patent Text Reader

Abstract

The present invention discloses a guidance optimization method and device for a telescopic mechanism, and relates to the technical field of machine learning. The method comprises: determining the multi-stage telescopic parts of the telescopic mechanism, collecting multi-stage structural parameters; sampling the distance of each stage of the telescopic parts, and obtaining a multi-stage sensor sample data set; using the structural parameters and the sample data set to train a multi-stage stable identification model; obtaining the target guidance distance; using the multi-stage telescopic parts guidance distance as a variable, calling the model for optimization, ending when the stability is met, and outputting the multi-stage optimized guidance distance; controlling the telescopic mechanism to be extended and retracted according to the multi-stage optimized guidance distance. The present application solves the technical problem that the existing telescopic mechanism relies on fixed structural parameters and control algorithms during the guidance process, is difficult to adapt to the guidance requirements under different working conditions, and is prone to deviations and instabilities during the guidance process, and realizes the effect of optimizing the guidance distance through a multi-stage stable identification model to improve the guidance stability and control accuracy of the telescopic mechanism.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, specifically to the field of machine learning technology, and in particular to a guidance optimization method and device for a telescopic mechanism. Background Art

[0002] With the continuous advancement of industrial manufacturing technology and the improvement of automation level, telescopic mechanisms play an increasingly important role in various mechanical equipment and production lines. The telescopic mechanism can achieve precise displacement and positioning, which is crucial to ensuring the stability of the equipment and production efficiency. However, with the continuous improvement of production requirements and the increasing complexity of application scenarios, the guidance accuracy and stability of the telescopic mechanism are also facing increasing challenges. Traditional telescopic mechanisms often rely on fixed structural parameters and control algorithms during the guidance process, and it is difficult to adapt to the guidance requirements under different working conditions. This limitation makes the telescopic mechanism prone to deviation and instability during the guidance process, which seriously affects the accuracy and reliability of the equipment. Summary of the invention

[0003] The embodiments of the present application provide a guidance optimization method and device for a telescopic mechanism, which solves the technical problem that the existing telescopic mechanism relies on fixed structural parameters and control algorithms during the guidance process, is difficult to adapt to the guidance requirements under different working conditions, and is prone to deviations and instability during the guidance process.

[0004] In view of the above problems, an embodiment of the present application provides a guidance optimization method and device for a telescopic mechanism.

[0005] A first aspect of an embodiment of the present application provides a guidance optimization method for a telescopic mechanism, the method comprising:

[0006] Determine a multi-stage telescopic component of the telescopic mechanism, and collect multi-stage structural parameters corresponding to the multi-stage telescopic component;

[0007] By sampling the distance of each level of telescopic parts, a multi-level sensing sample data set is obtained, wherein each level of sensing sample data set includes dynamic sensing data samples of each level of telescopic parts at different sample distances;

[0008] Training a multi-level stable recognition model according to the multi-level structural parameters and the multi-level sensing sample data set;

[0009] Obtaining a target guidance distance of the telescopic mechanism;

[0010] Taking the guide distance of the multi-stage telescopic member as the response variable, calling the multi-stage stability identification model to perform optimization with the target guide distance, and ending the optimization when the target guide stability is met, and outputting the multi-stage optimization guide distance corresponding to the multi-stage telescopic member one by one;

[0011] The telescopic mechanism is controlled to be telescopic according to the multi-stage optimal guiding distance.

[0012] A second aspect of the embodiments of the present application provides a guidance optimization device for a telescopic mechanism, the device comprising:

[0013] A parameter acquisition module, the parameter acquisition module is used to determine the multi-stage telescopic parts of the telescopic mechanism and to acquire multi-stage structural parameters corresponding to the multi-stage telescopic parts;

[0014] A distance sampling module, wherein the distance sampling module is used to obtain a multi-level sensing sample data set by performing distance sampling on each level of telescopic components, wherein each level of sensing sample data set includes dynamic sensing data samples of each level of telescopic components at different sample distances;

[0015] A model training module, the model training module is used to train a multi-level stable recognition model according to the multi-level structural parameters and the multi-level sensor sample data set;

[0016] A target-guided distance acquisition module, the target-guided distance acquisition module is used to acquire the target-guided distance of the telescopic mechanism;

[0017] An optimization module, wherein the optimization module is used to use the guide distance of the multi-stage telescopic member as a response variable, call the multi-stage stability identification model to perform optimization with the target guide distance, and terminate the optimization when the target guide stability is met, and output the multi-stage optimization guide distance corresponding to the multi-stage telescopic member one by one;

[0018] A telescopic control module is used to control the telescopic mechanism to telescope according to the multi-stage optimal guidance distance.

[0019] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0020] The above-mentioned guidance optimization method for a telescopic mechanism first determines the multi-stage telescopic parts of the telescopic mechanism and collects their respective structural parameters. Subsequently, a rich set of sensor data samples is obtained by sampling each stage of the telescopic parts at different distances. Afterwards, a multi-stage stable identification model is trained using these data sets and structural parameters. Then, the target guidance distance of the telescopic mechanism is obtained, and the multi-stage stable identification model is used to search for the best until a guidance distance that meets the stability requirements is found. Finally, according to these optimized guidance distances, the telescopic mechanism is accurately controlled to extend and retract. This method improves the guidance stability and control accuracy of the telescopic mechanism by collecting data, training models and optimizing guidance distances. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 A schematic flow chart of a guidance optimization method for a telescopic mechanism provided in an embodiment of the present application;

[0023] Figure 2 A schematic structural diagram of a guidance optimization device for a telescopic mechanism provided in an embodiment of the present application.

[0024] Explanation of the accompanying drawings: parameter acquisition module 1, distance sampling module 2, model training module 3, target-oriented distance acquisition module 4, optimization module 5, and telescopic control module 6. DETAILED DESCRIPTION

[0025] The embodiments of the present application provide a guidance optimization method and device for a telescopic mechanism to solve the technical problem that the existing telescopic mechanism relies on fixed structural parameters and control algorithms during the guidance process, is difficult to adapt to the guidance requirements under different working conditions, and is prone to deviations and instability during the guidance process.

[0026] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0027] It should be noted that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or are inherent to these processes, methods, products or devices.

[0028] Embodiment 1

[0029] like Figure 1 As shown, an embodiment of the present application provides a guidance optimization method for a telescopic mechanism, wherein the method comprises:

[0030] Determine a multi-stage telescopic component of the telescopic mechanism, and collect multi-stage structural parameters corresponding to the multi-stage telescopic component;

[0031] With the rapid development of industrial production and mechanical automation, telescopic mechanisms are increasingly used in various fields. As a key component for achieving precise displacement and positioning, the guiding accuracy and stability of the telescopic mechanism are directly related to the performance and production efficiency of the entire mechanical system. Therefore, optimizing the guiding performance of the telescopic mechanism is an inevitable requirement to meet the high-precision and high-efficiency requirements of modern industrial production.

[0032] In the present application embodiment, in the process of determining the guidance optimization of the telescopic mechanism, the system terminal first identifies and distinguishes the multi-stage telescopic parts in the telescopic mechanism. These telescopic parts have different characteristics due to their differences in structure and function, such as slender telescopic parts, sturdy telescopic parts, and sleeves as supports or connections. These different telescopic parts cooperate with each other to jointly realize the telescopic function of the telescopic mechanism. In order to gain a deeper understanding of the influence of each telescopic part on the guidance performance, the system terminal collects the multi-stage structural parameters corresponding to these multi-stage telescopic parts. These parameters not only reflect the basic information such as the size and shape of the telescopic parts, but also contain details such as material properties and connection methods that are closely related to the guidance performance. Through the collection and analysis of these multi-stage structural parameters, the guidance performance characteristics of the telescopic mechanism can be more accurately grasped, providing powerful data support for subsequent optimization work. In summary, through the identification of the multi-stage telescopic parts and the collection of the corresponding structural parameters, a solid foundation has been laid for the guidance optimization of the telescopic mechanism, and an important reference basis has been provided for subsequent precise control and performance improvement.

[0033] By sampling the distance of each level of telescopic parts, a multi-level sensing sample data set is obtained, wherein each level of sensing sample data set includes dynamic sensing data samples of each level of telescopic parts at different sample distances;

[0034] In one embodiment, in order to obtain the guidance performance data of the telescopic mechanism under different working conditions, the system terminal performs distance sampling on each level of telescopic parts. This process involves collecting dynamic sensor data samples of each level of telescopic parts at multiple different sample distances. These sensor data samples reflect the real-time status of the telescopic parts during the telescopic process, including key information such as position, speed, and acceleration. Through such sampling method, the system terminal constructs a rich multi-level sensor sample data set, which provides valuable data resources for subsequent guidance optimization.

[0035] Training a multi-level stable recognition model according to the multi-level structural parameters and the multi-level sensing sample data set;

[0036] In one embodiment, the system terminal divides the collected multi-level structural parameters and multi-level sensor sample data sets, and uses the structural parameters and sensor sample data sets of the same telescopic part as a sample training set. Subsequently, the structural parameters and sensor sample data sets in the divided sample training sets are respectively identified, and a stable recognition model is trained based on the identification results. This process is repeated until the structural parameters and sensor sample data sets of all telescopic parts are iterated, and the system terminal successfully obtains a multi-level stable recognition model. This model combines the structural characteristics of the telescopic part with its dynamic performance in the actual telescopic process through machine learning methods, thereby learning and identifying how the telescopic mechanism maintains stable guiding performance under different working conditions, helping the system terminal to better understand and optimize the performance of the telescopic mechanism.

[0037] Furthermore, training a multi-level stable recognition model according to the multi-level structural parameters and the multi-level sensing sample data set includes:

[0038] Acquire the first-level structural parameters and the first-level sensing sample data set corresponding to the first telescopic member, wherein the structural parameters include component connection parameters and component geometric parameters, and the sensing sample data set includes displacement sensing data and vibration sensing data;

[0039] Preferably, in order to construct a multi-level stable identification model, the system terminal first obtains relevant data of the first telescopic part from the multi-level structural parameters and the multi-level sensing sample data set. This includes the first-level structural parameters and the first-level sensing sample data set. The first-level structural parameters mainly describe the component connection method and geometric shape of the first telescopic part, for example, the connection method between the various components, the length and diameter of the telescopic part, etc. The first-level sensing sample data set records the displacement and vibration of the first telescopic part during the actual working process. These data reflect the dynamic performance of the telescopic part during the telescopic process. In summary, obtaining these detailed structural and motion data about the first telescopic part can provide important information support for subsequent analysis and optimization work.

[0040] Perform stability impact identification based on the first-level structural parameters and output a distance-structural stability impact index;

[0041] By identifying the distance-displacement deviation and the distance-vibration deviation of the first-level sensor sample data set, the distance-sensing stability is output;

[0042] Preferably, in order to gain a deeper understanding of the stability of the first telescopic member during the telescopic process, the system terminal first analyzes the first-level structural parameters of the first telescopic member, and deeply understands the influence mechanism of these parameters on the stability and guiding performance of the telescopic member. Subsequently, the stability of the first telescopic member at different telescopic distances is evaluated using professional knowledge. The stress distribution, deformation and bearing capacity of different structural parameters during the telescopic process are analyzed to determine their specific impact on stability. Specifically, the system terminal sets a series of specific indicators related to stability, such as stress peak value, deformation amount, bearing capacity loss rate, etc., according to the principles of engineering mechanics and mechanical design. For each indicator, the system terminal simulates the working state of the telescopic member at different telescopic distances through simulation components to quantify the specific values ​​of these indicators. Afterwards, the quantified indicators are combined with the corresponding telescopic distances to construct a distance-structure stability impact index, and output it. This index can reflect the comprehensive impact of structural parameters on the stability of the telescopic member at different telescopic distances, and provide an important basis for subsequent guidance optimization.

[0043] At the same time, the system terminal extracts displacement data and vibration data at different telescopic distances from the sensor sample data set, and sets the corresponding reference displacement value and standard vibration value for comparison with the actual data. Subsequently, the difference between the actual displacement and the theoretical displacement at each telescopic distance is calculated, that is, the displacement deviation, and the deviation between the actual vibration and the expected vibration at each telescopic distance is calculated, that is, the vibration deviation. After that, the change of the displacement deviation with the telescopic distance is analyzed based on the obtained displacement deviation, and the displacement stability is evaluated. The change trend of the vibration deviation with the telescopic distance is analyzed again, and the vibration stability is evaluated. Then, combined with historical experience, weights are assigned to the distance-displacement deviation and the distance-vibration deviation, that is, the importance of displacement and vibration is judged based on historical experience. Finally, according to the assigned weights, the system terminal performs weighted summation calculation on the distance-displacement deviation and the distance-vibration deviation, multiplies the distance-displacement deviation and the distance-vibration deviation with the corresponding weights, and then accumulates them to calculate a comprehensive index, that is, the distance-sensing stability. This step enables the system terminal to more intuitively understand the relationship between the displacement and vibration of the telescopic component and the telescopic distance during the telescopic process, thereby judging its sensing stability.

[0044] In summary, the system terminal used structural parameters and sensor data to conduct a comprehensive analysis of the stability of the first telescopic member and obtained corresponding stability indicators. These indicators provide important references for subsequent optimization work.

[0045] The feedback network layer is trained with the distance-structure stability influence index, and the recognition network layer is trained with the distance-sensing stability, and a first-level stable recognition model is output, and so on, a multi-level stable recognition model is output.

[0046] Preferably, after obtaining the two key indicators of distance-structural stability influence index and distance-sensing stability, the system terminal uses these indicators to train a multi-level stability recognition model so as to accurately identify the stability of the telescopic member. First, the system terminal uses the distance-structural stability influence index as part of the training data. These data include structural parameters and corresponding stability evaluation results at different telescopic distances. Then design the structure of the feedback network layer, determine the number of neurons in the input layer, hidden layer and output layer and the connection method between them. Subsequently, the feedback network layer is trained using the prepared training data. The output is calculated by forward propagation, compared with the true stability evaluation result, and the loss function is calculated. After that, the network weight is adjusted by the back propagation algorithm to minimize the loss function. Through iterative training, the network parameters are continuously adjusted until the feedback network layer achieves satisfactory performance on the verification set. During the training of the feedback network layer, the system terminal uses distance-sensing stability as another part of the training data. These data include sensor data and corresponding stability identification labels at different telescopic distances. Then use the same training method as above to train the designed recognition network layer to improve the performance of the recognition network layer. The role of the feedback network layer is to predict and evaluate the stability of the telescopic parts based on the structural parameters and telescopic distance. The role of the identification network layer is to identify the sensor stability of the telescopic parts under different working conditions by analyzing the sensor data. After completing the training of the feedback network layer and the identification network layer, the system terminal obtains the first-level stability identification model. This model can comprehensively consider the influence of structural parameters and sensor data to make a preliminary evaluation of the stability of the telescopic parts. Similarly, the system terminal continues to use the structural parameters and sensor data of other telescopic parts to train multi-level stability identification models. Due to the differences in telescopic parts, these models have their own characteristics in structure and capabilities. These different models can more comprehensively consider the various factors affecting the stability of the telescopic parts. Through the training and combination of multi-level models, more accurate and detailed stability identification results can be obtained.

[0047] Obtaining a target guidance distance of the telescopic mechanism;

[0048] In one embodiment, the system terminal analyzes the specific application scenarios and usage requirements of the telescopic mechanism. This involves a detailed analysis and understanding of the working environment, working conditions, and tasks to be completed of the telescopic mechanism. Subsequently, based on the application scenarios and usage requirements, the target position and state that the telescopic mechanism needs to achieve, i.e., the target guidance distance, is determined. The target guidance distance can be the maximum extension or shortening length of the telescopic mechanism, or a specific intermediate position. This distance will serve as an important basis for the subsequent optimization and control of the telescopic mechanism, helping to achieve precise guidance and stable operation of the telescopic mechanism.

[0049] Taking the guide distance of the multi-stage telescopic member as the response variable, calling the multi-stage stability identification model to perform optimization with the target guide distance, and ending the optimization when the target guide stability is met, and outputting the multi-stage optimization guide distance corresponding to the multi-stage telescopic member one by one;

[0050] In one embodiment, in finding the optimal guide distance combination for the stability performance of the multi-stage telescopic parts under different guide distances, the system terminal inputs the initial guide distance as an input variable into the multi-stage stability identification model, and the multi-stage stability identification model analyzes the input guide distance and outputs multi-stage stability, i.e., the response variable. Subsequently, the system terminal observes the change of the stability identification result with the guide distance through the multi-stage stability identification model. By continuously adjusting the guide distance, observe how the response variable output by the model changes. In this process, the system terminal performs weighted fusion on the response variable output each time, and compares the fusion result with the target-oriented stability to determine whether the fusion result reaches or exceeds the target-oriented stability. If the requirement of the target-oriented stability is not met, the system terminal controls the optimization direction of the next optimization round based on the penalty factor. If it is met, the guide distance under the current optimization round is output to obtain the multi-stage optimization guide distance corresponding to the multi-stage telescopic parts one by one, which will be used as the optimal guide distance used by the telescopic parts in practical applications to ensure its stability and performance. In summary, this process is to observe the change of the response variable output by the model by continuously adjusting the guide distance to find the guide distance combination that makes the telescopic parts stable.

[0051] Furthermore, calling the multi-level stable recognition model to perform optimization based on the target-oriented distance includes:

[0052] Using the multi-level stable identification model to stably identify the multi-level initial distance solution, and outputting multi-level stability;

[0053] Performing weighted fusion on the multi-level stability and outputting fused stability;

[0054] Preferably, the system terminal uses a multi-level stable identification model to identify the randomly selected multi-level initial distance solution, and the model evaluates the stability of each level according to the relationship between the learned structural parameters and sensor data and stability. The stability performance of each level of telescopic parts at different initial distances can be obtained, that is, multi-level stability. Subsequently, in order to obtain a more comprehensive and comprehensive stability evaluation result, the system terminal uses a weighted fusion method to process the multi-level stability. Weighted fusion means giving different weights according to the contribution of each level of telescopic parts to the overall stability, and then weighted averaging the stability of each level. In this way, the system terminal can integrate the stability information of different levels to obtain an indicator reflecting the overall stability, that is, fusion stability. In summary, this process uses a multi-level stable identification model to stably identify the multi-level initial distance solution and obtain the stability performance of each level of telescopic parts; then, the stability of each level is integrated by a weighted fusion method, and finally a fusion stability index is output to comprehensively evaluate the overall stability of the telescopic mechanism. This process is of great significance for optimizing the design of the telescopic mechanism and improving its working stability and reliability.

[0055] When the fusion stability does not satisfy the goal-oriented stability, an iterative penalty factor is set according to the size of the multi-level stability, and the optimization direction of the next optimization round is controlled based on the penalty factor until the goal-oriented stability is satisfied.

[0056] Preferably, when the fusion stability fails to achieve the target-oriented stability, the system terminal adjusts the optimization process according to the specific performance of the multi-level stability. Specifically, different iterative penalty factors are set according to the stability of each level of telescopic parts. These penalty factors are a weight adjustment method for telescopic parts with poor stability in the optimization process. In the next round of optimization, the system terminal uses these penalty factors to control the optimization direction. For telescopic parts with poor stability, their corresponding penalty factors will be increased so that they get more attention in the optimization process, thereby adjusting their guiding distance and improving their stability. On the contrary, for telescopic parts with better stability, their penalty factors will be reduced to maintain their current guiding distance. Through such iterative adjustments, the system terminal gradually optimizes the guiding distances of telescopic parts at all levels until the target-oriented stability is finally met. This process is a process of gradually approaching the goal. Through continuous trial and error and adjustment, a guiding distance combination that can achieve the overall stability target is found.

[0057] Furthermore, obtaining the multi-level initial distance solution includes:

[0058] Obtaining the multi-stage telescopic limit distance of the multi-stage telescopic member;

[0059] The multi-stage telescopic limit distance is used as a multi-stage solution space, and multi-stage initial distance solutions are randomly selected in the multi-stage solution space, wherein the multi-stage initial distance solutions include the initial solution of each stage of the multi-stage telescopic member.

[0060] Optionally, the system terminal obtains the telescopic limit distance of the multi-stage telescopic member at each stage, and these limit distances represent the maximum and minimum distances that each stage of the telescopic member can reach during the telescopic process. These telescopic limit distances are determined based on factors such as the structural design, material properties, and working conditions of the telescopic member, and are the range limits of the normal operation of the telescopic member. Subsequently, the system terminal uses these multi-stage telescopic limit distances as a multi-stage solution space, that is, defines the possible value range of the guide distance of each stage of the telescopic member during the optimization process. This solution space provides the system terminal with boundary conditions for searching for a suitable guide distance. Afterwards, in this multi-stage solution space, the system terminal generates a multi-stage initial distance solution by random selection. These initial solutions are the starting point of the optimization process, including the initial guide distance of each stage of the telescopic member in the multi-stage telescopic member. Through random selection, the system terminal can obtain a multi-stage initial solution as the starting point for subsequent optimization. In summary, randomly selecting a multi-stage initial distance solution in this solution space can provide basic data for subsequent stable identification and optimization processes. In this way, optimization can be started from multiple starting points, which increases the possibility of finding the optimal solution, and also provides more options for the optimization design of the telescopic mechanism.

[0061] Furthermore, the multi-level stable identification model is used to stably identify the multi-level initial distance solution and output the multi-level stability, including:

[0062] Establishing an input mapping relationship between the multi-level stable recognition model and the multi-level initial distance solution;

[0063] Based on the input mapping relationship, model matching is performed on the multi-level initial distance solutions respectively, and stable identification is performed according to the stable identification model obtained by matching, and the first-level stability, the second-level stability, ..., the N-th level stability are output, wherein N is the number of multi-level telescopic parts, and N is a positive integer greater than or equal to 1;

[0064] The first level stability, the second level stability, ..., the Nth level stability are output as the multi-level stability.

[0065] Optionally, after establishing a multi-level stable identification model, the system terminal establishes an input mapping relationship between these models and the multi-level initial distance solution, and clarifies which stable identification model each initial distance solution should correspond to for stability evaluation, that is, the telescopic component represented by each level of the initial distance solution corresponds to the stable identification model of the level in the multi-level stable identification model for the stability evaluation of the telescopic component. Subsequently, based on this input mapping relationship, the system terminal performs model matching on each level of the initial distance solution. The initial distance solution of each level is input into the stable identification model of the corresponding level. These models will perform stable identification on each level of the initial distance solution based on the relationship between the learned structural parameters and sensor data and stability. Through this process, the stability evaluation results of each level of the telescopic component under a given initial distance solution are obtained, that is, the first level stability, the second level stability, and up to the Nth level stability, where N is a positive integer greater than or equal to 1, representing the total number of multi-level telescopic components. The stability of each level is independently evaluated based on its corresponding stable identification model. Afterwards, the system terminal outputs the stability evaluation results of these levels as multi-level stability. This multi-level stability output contains the stability information of each level of telescopic components under a given initial distance solution, providing a comprehensive stability evaluation result for the system terminal.

[0066] Furthermore, the optimization direction of the next optimization round is controlled based on the penalty factor, including:

[0067] Obtaining a multi-stage distance solution of the multi-stage telescopic component in the tth round;

[0068] Calculating the proportion of the multi-level distance solution of the t-th round to the target-oriented distance, and generating a multi-level proportion coefficient of the t-th round;

[0069] Optionally, during the optimization process, the system terminal continuously obtains the multi-level distance solutions of the multi-level telescopic parts in different rounds. These distance solutions represent the current guidance distance of the telescopic parts at each level and are temporary results in the optimization process. For the multi-level distance solution of the tth round, the system terminal calculates its proportion relative to the target guidance distance. This proportion can reflect the closeness between the current distance solution and the target guidance distance, thereby helping the system terminal evaluate the quality of the current solution. The specific method of calculating the proportion is to divide the current distance solution of each level by the corresponding target guidance distance to obtain the proportion of each level. Subsequently, these proportions are combined to generate the multi-level proportion coefficient of the tth round. This coefficient is a comprehensive indicator that can reflect the overall closeness of the current multi-level distance solution. In summary, this process is to obtain the multi-level distance solution of the multi-level telescopic parts in the tth round, and calculate the proportion of these distance solutions relative to the target guidance distance, thereby generating a multi-level proportion coefficient. This coefficient can help evaluate the quality of the current solution and provide guidance for the subsequent optimization process.

[0070] Determining the t-th round multi-level weight of the multi-level telescopic component according to the t-th round multi-level proportion coefficient;

[0071] The multi-level weight of the t-th round is updated according to the multi-level penalty factor corresponding to the multi-level telescopic component to obtain the multi-level weight of the t+1-th round;

[0072] Optionally, during the optimization process, the system terminal determines the weight of the multi-level telescopic parts in the current round according to the multi-level proportion coefficient of the tth round. This proportion coefficient reflects the closeness of the current distance solution to the target-oriented distance, so it can be used to adjust the importance of each level of telescopic parts in the optimization process. Specifically, for telescopic parts with a larger proportion coefficient, since they are already closer to the target-oriented distance, the system terminal will assign a smaller weight or keep its weight unchanged to avoid excessive adjustment and stability degradation; while for telescopic parts with a smaller proportion coefficient, since they are farther away from the target-oriented distance, the system terminal will assign a larger weight so as to increase the adjustment in the next round of optimization and make it approach the target-oriented distance faster. Subsequently, the system terminal updates the multi-level weights of the tth round according to the multi-level penalty factors corresponding to the multi-level telescopic parts. The introduction of the penalty factor is to consider the stability performance of the telescopic parts at all levels in the previous optimization process. For telescopic parts with poor stability, the system terminal uses the penalty factor to further increase its weight when updating the weight, so that it gets more attention in the subsequent optimization; and for telescopic parts with better stability, the penalty factor is also used to maintain its current stability performance or give it a smaller weight during fine-tuning. Finally, the system terminal obtains the multi-level weights of the t+1th round, which will be used to guide the direction adjustment in the next round of optimization. By continuously iterating and updating the weights, the guidance distance of each level of telescopic parts can be gradually optimized until the target guidance stability is met.

[0073] The step length is optimized using the multi-level weights of the t+1th round, and a multi-level distance solution of the t+1th round is output.

[0074] Optionally, in the optimization process of the t+1th round, the system terminal uses the calculated multi-level weights to guide the step optimization process. These weights are determined based on the ratio of the distance solution of the previous round to the target-oriented distance, as well as the multi-level penalty factor, reflecting the importance and adjustment strength of each level of telescopic parts in the current optimization stage. In the step optimization, the telescopic parts with large weights represent that they are far away from the target-oriented distance and have poor stability, so they will be given a larger step adjustment to approach the target-oriented distance faster and improve stability. For telescopic parts with small weights, since they are closer to the target and have better stability, the step adjustment will be relatively small to maintain their current state or adjust within the fine-tuning range. Through such weight distribution, the system terminal can more specifically adjust the guidance distance of telescopic parts at all levels in the optimization process of the t+1th round to achieve the purpose of optimizing overall stability. Finally, according to the results of the step optimization, the system terminal outputs the multi-level distance solution of the t+1th round, which will be used as the starting point for the next round of optimization, and continue to iterate until the target-oriented stability is met.

[0075] Furthermore, the system terminal uses the multi-level penalty factor of the multi-level scaling component to adjust the multi-level weight of the t-th round to obtain the multi-level weight of the t+1-th round. The specific expression is as follows:

[0076] ;

[0077] in, Indicates the weight corresponding to the i-th level telescopic component in the t-th round. It indicates the weight corresponding to the i-th level retractable component in the t+1th round after the update. Represents the level of the elastic component, from 1 to N, where N is the total number of multi-level elastic components. t represents the iteration round, that is, the current optimization process is the first round. It represents the fusion stability of the multi-level distance solution of the tth round after being calculated by the multi-level stable recognition model. represents the target-oriented distance, that is, the distance solution corresponding to the desired stability. Divide by the sum of all level weights , and obtain the weight proportion of each level of the elastic component in the tth round. This step is to ensure that the sum of the weights of all levels is 1, which is convenient for subsequent calculations. Then, the weight of the i-th level elastic component is updated using the multi-level penalty factor. The penalty factor is obtained by Calculated, where e is the base of the natural logarithm. This penalty factor is (fusion stability) and (target guidance distance). The closer When the penalty factor is closer to 1, the weight changes less; when and When the difference is large, the penalty factor will deviate significantly from 1, resulting in a large change in the weight. Then, the weight of the i-th level elastic component in the tth round is multiplied by the penalty factor to obtain the weight of the i-th level elastic component in the t+1th round. Through this calculation process, the system terminal considers the weight distribution of the multi-level telescopic parts in the tth round and the deviation between the fusion stability and the target guidance distance, and updates the weights so as to more accurately adjust the guidance distance of the telescopic parts at all levels in the optimization process of the t+1th round, and finally achieve the optimization of the target guidance stability.

[0078] The telescopic mechanism is controlled to be telescopic according to the multi-stage optimal guiding distance.

[0079] In one embodiment, the process of the system terminal controlling the telescopic mechanism to telescope according to the multi-level optimal guidance distance is actually to guide the movement of the telescopic mechanism by using the calculated optimal guidance distance of the telescopic parts at each level. Specifically, the system terminal controls the telescopic movement of the telescopic mechanism according to the obtained optimal distance of the telescopic parts at each level, that is, the target position they should reach. In this process, the telescopic mechanism will gradually adjust the length of each level of telescopic parts according to the preset control logic and mechanism, so that it gradually approaches or reaches the calculated optimal distance. Such a control process can ensure that the telescopic mechanism reaches a relatively stable state as a whole, because the telescopic movement of each level is based on the calculated optimal solution. Through continuous optimization, the stability and performance of the telescopic mechanism can be gradually improved, so that it can better adapt to various working environments and task requirements, so as to achieve overall stability and performance improvement.

[0080] In summary, the embodiments of the present application have at least the following technical effects:

[0081] The embodiment of the present application determines the multi-stage telescopic parts of the telescopic mechanism and collects the structural parameters and sensor sample data sets of these telescopic parts. Subsequently, the multi-stage structural parameters and sensor sample data sets collected are used to train a multi-stage stability recognition model. This model can predict the stability performance of the telescopic parts based on a given telescopic distance. After that, the target-oriented distance of the telescopic mechanism is determined, and based on this, the trained multi-stage stability recognition model is used for optimization. During the optimization process, the multi-stage initial distance solution is stably identified, the multi-stage stability is output, and weighted fusion is performed accordingly to obtain the fusion stability. If the fusion stability does not meet the target-oriented stability, an iterative penalty factor will be set according to the size of the multi-stage stability, and the direction of the next round of optimization will be adjusted based on these factors until the target-oriented stability is met. In the optimization process, it also involves the acquisition of multi-stage initial distance solutions and the control of the optimization direction based on the penalty factor. When obtaining the initial distance solution, the telescopic limit of the telescopic part is comprehensively considered and randomly selected within the limit range. When controlling the optimization direction, the multi-level weights are updated according to the proportion of distance solutions and penalty factors in each round, and the step length optimization is performed based on this, gradually approaching the target guidance distance. Finally, according to the multi-level optimization guidance distance obtained by optimization, the telescopic mechanism is controlled to be telescopic to achieve overall stability and performance optimization. These technical effects jointly solve the technical problems that the existing telescopic mechanism relies on fixed structural parameters and control algorithms during the guidance process, which makes it difficult to adapt to the guidance requirements under different working conditions, resulting in deviations and instability in the guidance process. It achieves the effect of optimizing the guidance distance through a multi-level stable identification model and improving the guidance stability and control accuracy of the telescopic mechanism.

[0082] Embodiment 2

[0083] Based on the same inventive concept as the guidance optimization method of a telescopic mechanism in the aforementioned embodiment, Figure 2 As shown, the present application provides a guidance optimization device for a telescopic mechanism, and the device and method embodiments in the present application are based on the same inventive concept. The device includes:

[0084] A parameter acquisition module 1, wherein the parameter acquisition module 1 is used to determine the multi-stage telescopic parts of the telescopic mechanism and to acquire multi-stage structural parameters corresponding to the multi-stage telescopic parts;

[0085] A distance sampling module 2, wherein the distance sampling module 2 is used to obtain a multi-level sensing sample data set by performing distance sampling on each level of telescopic components, wherein each level of sensing sample data set includes dynamic sensing data samples of each level of telescopic components at different sample distances;

[0086] A model training module 3, wherein the model training module 3 is used to train a multi-level stable recognition model according to the multi-level structural parameters and the multi-level sensor sample data set;

[0087] A target-guided distance acquisition module 4, wherein the target-guided distance acquisition module 4 is used to acquire the target-guided distance of the telescopic mechanism;

[0088] An optimization module 5, wherein the optimization module 5 is used to use the guide distance of the multi-stage telescopic member as a response variable, call the multi-stage stability identification model to perform optimization with the target guide distance, and terminate the optimization when the target guide stability is met, and output the multi-stage optimization guide distance corresponding to the multi-stage telescopic member one by one;

[0089] The telescopic control module 6 is used to control the telescopic mechanism to telescope according to the multi-stage optimal guidance distance.

[0090] Furthermore, the model training module 3 is used to perform the following method:

[0091] Acquire the first-level structural parameters and the first-level sensing sample data set corresponding to the first telescopic member, wherein the structural parameters include component connection parameters and component geometric parameters, and the sensing sample data set includes displacement sensing data and vibration sensing data;

[0092] Perform stability impact identification based on the first-level structural parameters and output a distance-structural stability impact index;

[0093] By identifying the distance-displacement deviation and the distance-vibration deviation of the first-level sensor sample data set, the distance-sensing stability is output;

[0094] The feedback network layer is trained with the distance-structure stability influence index, and the recognition network layer is trained with the distance-sensing stability, and a first-level stable recognition model is output, and so on, a multi-level stable recognition model is output.

[0095] Furthermore, the optimization module 5 is used to execute the following method:

[0096] Using the multi-level stable identification model to stably identify the multi-level initial distance solution, and outputting multi-level stability;

[0097] Performing weighted fusion on the multi-level stability and outputting fused stability;

[0098] When the fusion stability does not satisfy the goal-oriented stability, an iterative penalty factor is set according to the size of the multi-level stability, and the optimization direction of the next optimization round is controlled based on the penalty factor until the goal-oriented stability is satisfied.

[0099] Furthermore, the optimization module 5 is used to execute the following method:

[0100] Obtaining the multi-stage telescopic limit distance of the multi-stage telescopic member;

[0101] The multi-stage telescopic limit distance is used as a multi-stage solution space, and multi-stage initial distance solutions are randomly selected in the multi-stage solution space, wherein the multi-stage initial distance solutions include the initial solution of each stage of the multi-stage telescopic member.

[0102] Furthermore, the optimization module 5 is used to execute the following method:

[0103] Establishing an input mapping relationship between the multi-level stable recognition model and the multi-level initial distance solution;

[0104] Based on the input mapping relationship, model matching is performed on the multi-level initial distance solutions respectively, and stable identification is performed according to the stable identification model obtained by matching, and the first-level stability, the second-level stability, ..., the N-th level stability are output, wherein N is the number of multi-level telescopic parts, and N is a positive integer greater than or equal to 1;

[0105] The first level stability, the second level stability, ..., the Nth level stability are output as the multi-level stability.

[0106] Furthermore, the optimization module 5 is used to execute the following method:

[0107] Obtaining a multi-stage distance solution of the multi-stage telescopic component in the tth round;

[0108] Calculating the proportion of the multi-level distance solution of the t-th round to the target-oriented distance, and generating a multi-level proportion coefficient of the t-th round;

[0109] Determining the t-th round multi-level weight of the multi-level telescopic component according to the t-th round multi-level proportion coefficient;

[0110] The multi-level weight of the t-th round is updated according to the multi-level penalty factor corresponding to the multi-level telescopic component to obtain the multi-level weight of the t+1-th round;

[0111] The step length is optimized using the multi-level weights of the t+1th round, and a multi-level distance solution of the t+1th round is output.

[0112] Furthermore, the optimization module 5 is used to execute the following method:

[0113] The multi-level weights of the tth round are updated according to the multi-level penalty factors corresponding to the multi-level scaling components, and the expression is as follows:

[0114] ;

[0115] , is the number of N multi-stage telescoping parts, t is the number of iterations, Characterizes the multi-level weight of the tth round corresponding to the multi-level telescopic parts in the tth round, Represents the multi-level weight of the t+1th round corresponding to the multi-level telescopic component in the t+1th round, In order to use the multi-level stable recognition model to calculate the fusion stability of the multi-level distance solution of the tth round, The target-oriented distance.

[0116] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0117] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0118] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A guidance optimization method for a telescopic mechanism, characterized in that: The method comprises: Determine the multi-stage telescopic parts of the telescopic mechanism, and collect the multi-stage structural parameters corresponding to the multi-stage telescopic parts, wherein the multi-stage structural parameters include the size and shape of the telescopic parts, and the material properties and connection methods that are closely related to the guiding performance; By sampling the distance of each level of telescopic parts, a multi-level sensing sample data set is obtained, wherein each level of sensing sample data set includes dynamic sensing data samples of each level of telescopic parts at different sample distances; Training a multi-level stable recognition model according to the multi-level structural parameters and the multi-level sensing sample data set; Obtaining a target guidance distance of the telescopic mechanism; Taking the guide distance of the multi-stage telescopic member as the response variable, calling the multi-stage stability identification model to perform optimization with the target guide distance, and ending the optimization when the target guide stability is met, and outputting the multi-stage optimization guide distance corresponding to the multi-stage telescopic member one by one; The telescopic mechanism is controlled to be telescopic according to the multi-stage optimal guiding distance.

2. The method according to claim 1, characterized in that Calling the multi-level stable recognition model to perform optimization based on the target-oriented distance includes: Using the multi-level stable identification model to stably identify the multi-level initial distance solution, and outputting multi-level stability; Performing weighted fusion on the multi-level stability and outputting fused stability; When the fusion stability does not satisfy the goal-oriented stability, an iterative penalty factor is set according to the size of the multi-level stability, and the optimization direction of the next optimization round is controlled based on the penalty factor until the goal-oriented stability is satisfied.

3. The method according to claim 2, characterized in that Obtaining the multi-level initial distance solution includes: Obtaining the multi-stage telescopic limit distance of the multi-stage telescopic member; The multi-stage telescopic limit distance is used as a multi-stage solution space, and multi-stage initial distance solutions are randomly selected in the multi-stage solution space, wherein the multi-stage initial distance solutions include the initial solution of each stage of the multi-stage telescopic member.

4. The method according to claim 1, characterized in that According to the multi-level structural parameters and the multi-level sensing sample data set, a multi-level stable recognition model is trained, including: Acquire the first-level structural parameters and the first-level sensor sample data set corresponding to the first telescopic member, wherein the structural parameters include component connection parameters and component geometric parameters, and the sensor sample data set includes displacement sensor data and vibration sensor data; Perform stability impact identification based on the first-level structural parameters and output a distance-structural stability impact index; By identifying the distance-displacement deviation and the distance-vibration deviation of the first-level sensor sample data set, the distance-sensing stability is output; The feedback network layer is trained with the distance-structure stability influence index, and the recognition network layer is trained with the distance-sensing stability, and a first-level stable recognition model is output, and so on, a multi-level stable recognition model is output.

5. The method according to claim 2, characterized in that The multi-level stable identification model is used to stably identify the multi-level initial distance solution and output multi-level stability, including: Establishing an input mapping relationship between the multi-level stable recognition model and the multi-level initial distance solution; Based on the input mapping relationship, model matching is performed on the multi-level initial distance solutions respectively, and stable identification is performed according to the stable identification model obtained by matching, and the first-level stability, the second-level stability, ..., the N-th level stability are output, wherein N is the number of multi-level telescopic parts, and N is a positive integer greater than or equal to 1; The first level stability, the second level stability, ..., the Nth level stability are output as the multi-level stability.

6. The method according to claim 2, characterized in that The optimization direction of the next optimization round is controlled based on the penalty factor, and the method includes: Obtaining a multi-stage distance solution of the multi-stage telescopic component in the tth round; Calculating the proportion of the multi-level distance solution of the t-th round to the target-oriented distance, and generating a multi-level proportion coefficient of the t-th round; Determining the t-th round multi-level weight of the multi-level telescopic component according to the t-th round multi-level proportion coefficient; The multi-level weight of the t-th round is updated according to the multi-level penalty factor corresponding to the multi-level telescopic component to obtain the multi-level weight of the t+1-th round; The step length is optimized using the multi-level weights of the t+1th round, and the multi-level distance solution of the t+1th round is output.

7. The method according to claim 6, characterized in that The multi-level weights of the tth round are updated according to the multi-level penalty factors corresponding to the multi-level scaling components, and the expression is as follows: i=1,2,...,N, N is the number of N multi-stage telescoping components, t is the number of iterations, D t (i represents the multi-level weight of the t-th round corresponding to the multi-level telescopic parts in the t-th round, D t+1 (i) represents the multi-level weight of the t+1th round corresponding to the multi-level telescopic component in the t+1th round, θ t In order to use the multi-level stable recognition model to calculate the fusion stability of the multi-level distance solution of the tth round, The target-oriented distance.

8. A guiding optimization device for a telescopic mechanism, characterized in that: A method for optimizing the guidance of a telescopic mechanism according to any one of claims 1 to 7, the device comprising: Parameter collection module: determining the multi-stage telescopic parts of the telescopic mechanism, and collecting the multi-stage structural parameters corresponding to the multi-stage telescopic parts, wherein the multi-stage structural parameters include the size and shape of the telescopic parts, and the material properties and connection methods that are closely related to the guiding performance; Distance sampling module: by sampling the distance of each level of telescopic parts, a multi-level sensing sample data set is obtained, wherein each level of sensing sample data set includes dynamic sensing data samples of each level of telescopic parts at different sample distances; Model training module: training a multi-level stable recognition model according to the multi-level structural parameters and the multi-level sensing sample data set; A target-guided distance acquisition module is used to acquire the target-guided distance of the telescopic mechanism; Optimization module: taking the guide distance of the multi-stage telescopic member as the response variable, calling the multi-stage stability identification model to perform optimization with the target guide distance, and ending the optimization when the target guide stability is met, and outputting the multi-stage optimization guide distance corresponding to the multi-stage telescopic member one by one; Telescopic control module: controls the telescopic mechanism to telescope according to the multi-level optimal guidance distance.