A method, device and robot for constructing an industrial robot model

Abstract

The present invention provides a method, device and robot for constructing an industrial robot model, and relates to the technical field of robot performance optimization. The present invention effectively screens out key performance indicators that are highly correlated with equipment operation status data by setting a screening threshold of an associated screening coefficient, thereby improving the accuracy and robustness of the model; adopts a formulaic initial model construction method to simplify the model construction process and improve efficiency; finally, by calculating the ratio of associated screening coefficients in a subsequent time period and dynamically adjusting the screening threshold, adaptive optimization of model parameters is achieved, and the adaptability and intelligence level of the model in a complex production environment are enhanced; unnecessary complex calculations can be reduced, and the technical efficacy of rapid model generation is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of robot performance optimization, and in particular to a method and device for constructing an industrial robot model and a robot. Background Art

[0002] The application of industrial robots in the manufacturing and automation fields is constantly expanding, and their technology is also developing rapidly. Traditional industrial robot models usually rely on preset motion paths and fixed task parameters, and are difficult to flexibly adapt to dynamically changing production environments. With the advancement of artificial intelligence and big data technologies, data-driven model building methods have gradually become mainstream. By collecting and analyzing equipment operation status data in real time, these methods can dynamically adjust the robot's operation strategy and improve its adaptability and efficiency. However, most of the current model building methods still face problems such as complex data processing and inaccurate screening of key indicators, which restricts the intelligent application of industrial robots in complex tasks.

[0003] In the prior art, the publication number is CN113043273A, and the name is a method and system for constructing an industrial robot model. The method for constructing an industrial robot model includes: a model parameter input step: designing a human-computer interaction interface, and inputting model parameters in the human-computer interaction interface; a model parameter optimization step: optimizing the model parameters through a genetic algorithm, and outputting the optimization result parameters; a model generation step: obtaining the transmission part selection result parameters through the optimization result parameters, and generating the industrial robot model through a secondary developed Pro / TOOLKI T program based on the transmission part selection result parameters.

[0004] The shortcomings of existing technologies are mainly reflected in the lack of in-depth mining of operating status data, the imperfect screening and sorting mechanism of key performance indicators, and the difficulty of existing methods to effectively identify indicators that are highly correlated with equipment performance, resulting in insufficient accuracy and robustness of the model. In addition, the dynamic adjustment capability of existing models is limited, and it is impossible to adaptively optimize parameters according to changes in operating status in different time periods. Therefore, in the process of building industrial robot models, there is an urgent need for an innovative method that can accurately screen key performance indicators and dynamically optimize according to actual operating status;

[0005] The above information disclosed in the above Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The object of the present invention is to provide a method, device and robot for constructing an industrial robot model to solve the problems raised in the above background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A method for constructing an industrial robot model, the specific steps comprising:

[0009] Step S1: Collecting equipment operation status data and initial performance indicators of the industrial robot in the current time period T1;

[0010] Step S2: Obtain the equipment operation status data and initial performance indicators of the current time period T1 for analysis, generate a correlation screening coefficient, and set a screening threshold of the correlation screening coefficient. The screening threshold is used to screen out a key indicator set that is highly correlated with the equipment operation status data from the initial performance indicators, and sort the various performance indicators in the key indicator set according to the degree of correlation;

[0011] Step S3: Filter out the top three key performance indicators with the highest correlation with the equipment operation status data from the sorted key indicator set;

[0012] Step S4: Receive and analyze the key performance indicators obtained in step S3, generate an initial model, and optimize the parameters of the initial model to obtain an industrial robot model:

[0013] Step S5: In the next time period T2 adjacent to the current time period, the screening threshold of the current time period T1 is used as the preset threshold of the next time period T2, and the ratio between the corresponding associated screening coefficients of the two adjacent time periods T1 and T2 is calculated. Based on the ratio, analysis is performed to determine whether it is necessary to optimize and adjust the preset threshold of the next time period T2.

[0014] Furthermore, in the next time period T2 immediately adjacent to the current time period, the screening threshold of the current time period T1 is used as the preset threshold of the next time period T2, and the ratio between the corresponding associated screening coefficients of the two adjacent time periods T1 and T2 is calculated. Based on the ratio, an analysis is performed to determine whether it is necessary to optimize and adjust the screening threshold of the next time period T2, specifically including:

[0015] Calculate and mark the correlation screening coefficient of the next time period T2 as Cρ2, and calculate the ratio between Cρ and Cρ2 to obtain μ1 is the adjustment coefficient. The setting of μ1 is used to make The value range is limited to (0,1); and based on experimental demonstration or expert group analysis, the judgment threshold of the ratio is determined to be R1, R1∈(0,1);

[0016] when When , the adjustment formula for C′ρ is:

[0017]

[0018] when When , the adjustment formula for C′ρ is:

[0019]

[0020] Wherein, C″ρ is the preset threshold after optimization adjustment.

[0021] A device for constructing an industrial robot model, the device being used to execute a method for constructing the industrial robot model, comprising:

[0022] Data acquisition module: used to collect the equipment operation status data and initial performance indicators of the industrial robot in the current time period T1;

[0023] Key indicator set generation module: used to obtain the equipment operation status data and initial performance indicators of the current time period T1 for analysis, generate a correlation screening coefficient, and set a screening threshold of the correlation screening coefficient. The screening threshold is used to screen out a key indicator set that is highly correlated with the equipment operation status data from the initial performance indicators, and sort the various performance indicators in the key indicator set according to the degree of correlation;

[0024] Screening module: used to screen out the top three key performance indicators with the highest correlation with the equipment operation status data from the sorted key indicator set;

[0025] Initial model generation module: receives and analyzes the key performance indicators, generates an initial model, and optimizes the parameters of the initial model to obtain an industrial robot model;

[0026] Optimization and adjustment module: It is used to use the screening threshold of the current time period T1 as the preset threshold of the next time period T2 in the next time period T2 immediately adjacent to the current time period, calculate the ratio between the corresponding associated screening coefficients of the two adjacent time periods T1 and T2, and analyze based on the ratio to determine whether it is necessary to optimize and adjust the screening threshold in the subsequent time period, so as to adaptively optimize the operating status parameters of the industrial robot model in different time periods.

[0027] A robot comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the method for constructing an industrial robot model are implemented.

[0028] Compared with the prior art, the beneficial effects of the present invention are as follows: by setting the screening threshold of the correlation screening coefficient, the key performance indicators that are highly correlated with the equipment operation status data are effectively screened out, thereby improving the accuracy and robustness of the model; secondly, by sorting the key performance indicators and selecting the top three indicators, the accuracy and effectiveness of the model are further improved;

[0029] In addition, the use of a formulated initial model construction method simplifies the model construction process and improves efficiency. Finally, by calculating the ratio of the associated screening coefficients in the subsequent time period and dynamically adjusting the screening threshold, the adaptive optimization of the model parameters is achieved, enhancing the adaptability and intelligence of the model in complex production environments. It can reduce unnecessary complex calculations and achieve the technical efficacy of rapid model generation. These improvements not only improve the performance of industrial robots in dynamic environments, but also enhance the breadth and practicality of their applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of the overall method flow of the present invention;

[0031] Figure 2 It is a block diagram of the state device module of the present invention. DETAILED DESCRIPTION

[0032] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0033] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0034] Embodiment 1:

[0035] See also Figure 1 , the present invention provides a technical solution:

[0036] A method for constructing an industrial robot model, the specific steps comprising:

[0037] Step S1: Collecting equipment operation status data and initial performance indicators of the industrial robot in the current time period T1;

[0038] Step S2: Obtain the equipment operation status data and initial performance indicators of the current time period T1 for analysis, generate a correlation screening coefficient, and set a screening threshold of the correlation screening coefficient. The screening threshold is used to screen out a key indicator set that is highly correlated with the equipment operation status data from the initial performance indicators, and sort the various performance indicators in the key indicator set according to the degree of correlation;

[0039] Step S3: Filter out the top three key performance indicators with the highest correlation with the equipment operation status data from the sorted key indicator set;

[0040] Step S4: Receive and analyze the key performance indicators obtained in step S3, generate an initial model, and optimize the parameters of the initial model to obtain an industrial robot model:

[0041] Step S5: In the subsequent time period T2 immediately adjacent to the current time period, the screening threshold of the current time period T1 is used as the preset threshold of the subsequent time period T2, and the ratio between the corresponding associated screening coefficients of the two adjacent time periods T1 and T2 is calculated. Based on the ratio, analysis is performed to determine whether it is necessary to optimize and adjust the preset threshold of the subsequent time period T2, so as to adaptively optimize the operating status parameters of the industrial robot model in different time periods.

[0042] To further explain, the initial performance indicators include accuracy, speed, load capacity, energy consumption, and lifespan;

[0043] Equipment operating status data includes operating time, environmental conditions, and load changes;

[0044] Set the number of acquisitions in the current time period T1 to n, and obtain the acquisition number sequence {1, 2, ..., i, ..., n}, where i represents the index of the i-th acquisition number, and n represents the total number of acquisitions in the current time period;

[0045] The generation of the correlation screening coefficient includes:

[0046] Acquire equipment operation status data for analysis, and generate an operation status evaluation coefficient for graded evaluation of the operation status of the industrial robot;

[0047] Acquire initial performance indicators for analysis, and generate performance status evaluation coefficients for grading and evaluating initial performance status of the industrial robot;

[0048] The operating status evaluation coefficient and the performance status evaluation coefficient are combined and analyzed to generate a correlation screening coefficient.

[0049] Further explanation: The initial performance indicators and equipment operating status data are explained as follows:

[0050] The accuracy is recorded as Jd. The accuracy Jd is quantified by measuring the operation deviation displacement between the actual operation of the robot and the preset target value and taking the reciprocal; the calculation formula is as follows:

[0051]

[0052] Among them, Pc is the operation deviation displacement between the actual operation of the robot and the target value;

[0053] The speed is recorded as Sd, and the speed Sd is quantified by the average movement speed of the robot in the current time period T1;

[0054] The load capacity is denoted as Fz, which is quantified by the maximum load mass that the robot can bear;

[0055] The energy consumption is recorded as Nh, which is quantified by the average energy consumption of the robot in the current time period T1;

[0056] The life span is denoted as Sm, which is quantified by the expected service life of the robot under the preset working environment;

[0057] The running time is recorded as Yt, and the running time Yt is quantified by the actual working time of the robot in the current time period T1;

[0058] The environmental condition is recorded as Hj, which includes the temperature, humidity and vibration amplitude factors of the installation part of the industrial robot. Hj is comprehensively expressed by the following calculation formula:

[0059] Hj=w T WT+w H SH+w V ·ZV

[0060] Among them, WT, SH, and ZV represent the normalized values ​​of temperature, humidity, and vibration amplitude, respectively. T 、w H 、w V is the weight coefficient of the corresponding factor;

[0061] The load change is denoted as F l, and the load change F l is quantified by the load change rate of the robot in the current time period T1.

[0062] Further explanation: the construction of the operating status evaluation coefficient specifically includes:

[0063] Normalize the equipment operation status data collected in the current time period T1, including the operation time, environmental conditions, and load changes, to ensure the consistency of different data dimensions;

[0064] Use a fractional function to generate the operation status evaluation coefficient R(Yt, Hj, Fl), ensuring that its value range is (0, 1):

[0065]

[0066] Among them, w t 、w e 、w l are the weight coefficients of running time, environmental conditions, and load change respectively; These weight coefficients are set through empirical data or an expert system;

[0067] According to the output value of the operation status evaluation coefficient R(Yt, Hj, Fl), conduct a grading evaluation of the operation status of the industrial robot into four levels: excellent, good, medium, and poor. The specific grading method is as follows:

[0068] When R(Yt, Hj, Fl) > 0.75, the operation status of the industrial robot is at the excellent level; the excellent level indicates that the compliance of each parameter of the operation status of the industrial robot with the corresponding preset value is above 90%; the corresponding preset value is determined by the industrial robot manufacturer or the expert group through experiments and will not be elaborated later;

[0069] When 0.5 < R(Yt, Hj, Fl) ≤ 0.75, the operation status of the industrial robot is at the good level; the good level indicates that the compliance of each parameter of the operation status of the industrial robot with the corresponding preset value is between 75% and 90%, excluding 75%;

[0070] When 0.25 < R(Yt, Hj, Fl) ≤ 0.5, the operation status of the industrial robot is at the medium level; the medium level indicates that the compliance of each parameter of the operation status of the industrial robot with the corresponding preset value is between 45% and 75%, excluding 45%;

[0071] When R(Yt, Hj, Fl) ≤ 0.25, the operation status of the industrial robot is at the poor level. The poor level indicates that the compliance of each parameter of the operation status of the industrial robot with the corresponding preset value is below 45%.

[0072] Further explanation, the construction of the performance status evaluation coefficient specifically includes:

[0073] Normalize the initial performance indicators including accuracy, speed, load capacity, energy consumption, and lifespan collected during the current time period T1 to ensure the consistency of different data dimensions;

[0074] Use a fractional function to generate the performance status evaluation coefficient P(Jd, Sd, Fz, Nh, Sm), ensuring that its value range is (0, 1):

[0075]

[0076] Among them, w Jd , w Sd , w Fz , w Nh , w Sm are the weight coefficients of precision, speed, load capacity, energy consumption, and lifespan respectively; w Jd , w Sd , w Fz , w Nh , w Sm The weight coefficients are set through empirical data or an expert system;

[0077] η8 is a correction term. η8 combines the adjustment of the weight coefficients w Jd , w Sd , w Fz , w Nh , w Sm to ensure the correct classification of the following four levels: high, medium-high, medium, and low;

[0078] Based on the output value of the performance status evaluation coefficient P(Jd, Sd, Fz, Nh, Sm), the initial performance status of the industrial robot is evaluated at four levels: high, medium-high, medium, and low; The specific classification method is as follows:

[0079] When P(Jd, Sd, Fz, Nh, Sm) > 0.81, the initial performance status of the industrial robot is at a high level; A high level indicates that the compliance of each parameter of the initial performance status of the industrial robot with the corresponding preset value is above 85%;

[0080] When 0.65 < P(Jd, Sd, Fz, Nh, Sm) ≤ 0.81, the initial performance status of the industrial robot is at a medium-high level; A medium-high level indicates that the compliance of each parameter of the initial performance status of the industrial robot with the corresponding preset value is between 72% and 85%, excluding 72%;

[0081] When 0.36 < P(Jd, Sd, Fz, Nh, Sm) ≤ 0.65, the initial performance status of the industrial robot is at a medium level; A medium level indicates that the compliance of each parameter of the initial performance status of the industrial robot with the corresponding preset value is between 41% and 72%, excluding 41%;

[0082] When P(Jd, Sd, Fz, Nh, Sm) ≤ 0.36, the initial performance status of the industrial robot is at a low level. A low level indicates that the compliance of each parameter of the initial performance status of the industrial robot with the corresponding preset value is below 41%.

[0083] To further illustrate, the operating status evaluation coefficient R (Yt, Hj, Fl) and the performance status evaluation coefficient P (Jd, Sd, Fz, Nh, Sm) are combined and analyzed to calculate the correlation screening coefficient Cρ between the two:

[0084]

[0085] Among them, L1 R and L2 P They are the positive adjustment coefficients of R(Yt,Hj,Fl) and P(Jd,Sd,Fz,Nh,Sm), 0.12≤L1 R ≤0.88, 0.11≤L2 p ≤0.92; in this embodiment, L1 R and L2 P The average value is 0.5;

[0086] Set the screening threshold of the correlation screening coefficient Cρ to C′ρ, C′ρ is a positive number less than 1. When the correlation screening coefficient Cρ is greater than C′ρ, The accuracy, speed, load capacity, energy consumption and life indicators are taken as the key indicator set highly related to the equipment operation status data; where η1, η2, η3, η4 and η5 are the adjustment coefficients of the corresponding performance indicators respectively;

[0087] w Jd ≥C′ρ+η1 indicates that the corresponding accuracy is a key performance indicator that is highly correlated with the equipment operation status data; The description of the rest of the contents is the same as the above description and will not be repeated here;

[0088] The specific values ​​of η1, η2, η3, η4, and η5 are obtained based on experimental demonstration or systematic analysis by an expert group;

[0089] The various performance indicators in the key indicator set are sorted from large to small according to the degree of correlation; the top three key performance indicators are represented by X1, X2, and X3 respectively.

[0090] Further, receiving the generated initial model and performing model construction to generate the industrial robot model specifically includes:

[0091] a1, a2, and a3 are determined based on historical data or expert experience; a1, a2, and a3 are all positive numbers, and a1+a2+a3=1; a1, a2, and a3 values ​​are all within the range of 0 to 1;

[0092] The weight coefficient of the initial model is optimized using an optimization algorithm, and the optimization algorithm uses a gradient descent algorithm;

[0093] The updated and optimized values ​​of the initial weight coefficients a1, a2, and a3 are recorded as a1′, a2′, and a3′ respectively; the specific steps of the gradient descent algorithm include:

[0094] 1.1) Initialize the weight coefficients a1, a2, and a3 to the same value;

[0095] 1.2) Define the loss function Loss to measure the error of the model:

[0096]

[0097] Among them, P total,i =a1×X1 i +a2×X2 i +a3×X3 i ;P total,i represents the performance index value collected for the i-th time; P measured,i represents the actual measurement value collected at the i-th time;

[0098] 1.3) Calculate the gradient of the loss function with respect to each weight coefficient:

[0099]

[0100] in, Represents the partial derivative of the loss function Loss with respect to the variable a1;

[0101] Represents the partial derivative of the loss function Loss with respect to the variable a2;

[0102] Represents the partial derivative of the loss function Loss with respect to the variable a3;

[0103] 1.4) Update the weight coefficient according to the learning rate α:

[0104]

[0105] Among them, a1′, a2′, and a3′ are all updated and optimized weight coefficients;

[0106] 1.5) Repeat steps 1.2) to 1.4) and continue to iteratively update the weight coefficients until the loss function converges or reaches a predetermined number of iterations;

[0107] Use the validation data set to validate the optimized model and evaluate the model's predictive performance;

[0108] Calculate the mean square error (MSE) as a performance evaluation indicator:

[0109]

[0110] Apply the optimized weight coefficients a1′, a2′, and a3′ to the industrial robot model to form the final industrial robot model. The specific construction steps include:

[0111] According to the formula P total,n =a1×X1 n +a2×X2 n +a3×X3 n Calculate the total performance index of the industrial robot at the current time, and adjust the operating parameters and control strategy of the industrial robot according to the calculated total performance index to achieve optimal performance;

[0112] The detailed steps are:

[0113] After the optimized weight coefficients are applied to the robot model, the model is verified;

[0114] Validation set testing: Use a validation dataset that was not used in training to test the model to evaluate its performance and accuracy.

[0115] Cross-validation: Through multiple cross-validations, the performance of the model on different data sets is evaluated to ensure its generalization ability.

[0116] The optimized model is integrated into the robot system so that it can work in coordination with other system modules (such as sensors, actuators, controllers, etc.).

[0117] Software Integration: Convert the model into a format suitable for the robot control system and integrate it into the existing software framework.

[0118] Hardware integration: Ensure that the model can communicate with the robot hardware in real time and control the robot to perform tasks.

[0119] The integrated system is fully tested to ensure that all modules work together and the system functions and performance meet expectations.

[0120] Unit testing: Test each independent module to ensure that it runs normally individually.

[0121] Integration testing: Test the overall functionality of the system and verify that all modules work together.

[0122] Field testing: Testing the robot’s performance in a real or simulated working environment to ensure it can operate reliably in real applications.

[0123] Through test feedback, the model and system are further optimized and debugged.

[0124] Model tuning: Based on the test results, the model needs to be fine-tuned to improve its performance.

[0125] System debugging: Resolve any software or hardware issues found during testing to ensure system stability.

[0126] Based on selected X1 n , X2 n 、X3 n The value range is used to determine P total,n The value range of is (M1, M2) M1<M2;

[0127] When P total,n The closer the value is to M2, the better the comprehensive performance of the industrial robot under the current conditions. total,n The judgment threshold is Q1; Q1 is between M1 and M2;

[0128] When P total,n ≥Q1, it means the optimization effect is good; when P total,n When <Q1, it indicates poor optimization effect;

[0129] High speed and precision X1n: The robot has high movement speed and positioning accuracy, meeting the needs of high-precision manufacturing and fast operation.

[0130] High stability and load capacity; X2n: The robot is very stable during operation and can withstand large loads, making it suitable for heavy-load scenarios;

[0131] Low energy consumption and high efficiency X3n: The robot consumes less energy during operation and has high overall energy utilization efficiency, which helps save operating costs;

[0132] When P total,n The closer the value is to M1, the worse the overall performance of the industrial robot is, and optimization and adjustment are needed;

[0133] Further explanation: In the next time period T2 immediately adjacent to the current time period, the screening threshold of the current time period T1 is used as the preset threshold of the next time period T2, and the ratio between the corresponding associated screening coefficients of the two adjacent time periods T1 and T2 is calculated. Based on the ratio, analysis is performed to determine whether it is necessary to optimize and adjust the screening threshold of the next time period T2, specifically including:

[0134] Calculate and mark the correlation screening coefficient of the next time period T2 as Cρ2, and calculate the ratio between Cρ and Cρ2 to obtain μ1 is the adjustment coefficient. The setting of μ1 is used to make The value range is limited to (0,1); and based on experimental demonstration or expert group analysis, the judgment threshold of the ratio is determined to be R1, R1∈(0,1);

[0135] when When , it means that the correlation screening coefficient of T1 and T2 shows a downward trend, and the adjustment formula for C′ρ is:

[0136]

[0137] when When , it means that the correlation screening coefficient of T1 and T2 shows an upward trend, and the adjustment formula for C′ρ is:

[0138]

[0139] Wherein, C″ρ is the preset threshold after optimization adjustment.

[0140] In summary, it is possible to reduce unnecessary complex calculations and achieve the technical effect of rapidly generating models.

[0141] Embodiment 2:

[0142] See also Figure 2 , a device for constructing an industrial robot model, the device being used to execute the method for constructing the industrial robot model, comprising:

[0143] Data acquisition module: used to collect the equipment operation status data and initial performance indicators of the industrial robot in the current time period T1;

[0144] Key indicator set generation module: used to obtain the equipment operation status data and initial performance indicators of the current time period T1 for analysis, generate a correlation screening coefficient, and set a screening threshold of the correlation screening coefficient. The screening threshold is used to screen out a key indicator set that is highly correlated with the equipment operation status data from the initial performance indicators, and sort the various performance indicators in the key indicator set according to the degree of correlation;

[0145] Screening module: used to screen out the top three key performance indicators with the highest correlation with the equipment operation status data from the sorted key indicator set;

[0146] Initial model generation module: receives and analyzes the key performance indicators, generates an initial model, and optimizes the parameters of the initial model to obtain an industrial robot model;

[0147] Optimization and adjustment module: It is used to use the screening threshold of the current time period T1 as the preset threshold of the next time period T2 in the next time period T2 immediately adjacent to the current time period, calculate the ratio between the corresponding associated screening coefficients of the two adjacent time periods T1 and T2, and analyze based on the ratio to determine whether it is necessary to optimize and adjust the screening threshold in the subsequent time period, so as to adaptively optimize the operating status parameters of the industrial robot model in different time periods.

[0148] Embodiment three:

[0149] A robot comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the method for constructing an industrial robot model are implemented.

[0150] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0151] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0152] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0153] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for constructing an industrial robot model, characterized in that: The specific steps include: Step S1: Collecting equipment operation status data and initial performance indicators of the industrial robot in the current time period T1; Initial performance indicators include accuracy, speed, load capacity, energy consumption, and lifespan; The accuracy is recorded as Jd, the speed is recorded as Sd, the load capacity is recorded as Fz, the energy consumption is recorded as Nh, and the life is recorded as Sm; Equipment operating status data includes operating time, environmental conditions, and load changes; Let the running time be Yt, the environmental condition be Hj, and the load change be Fl; Step S2: Obtain the equipment operation status data and initial performance indicators of the current time period T1 for analysis, generate a correlation screening coefficient, and set a screening threshold of the correlation screening coefficient. The screening threshold is used to screen out a key indicator set that is highly correlated with the equipment operation status data from the initial performance indicators, and sort the various performance indicators in the key indicator set according to the degree of correlation; Use the fractional function to generate the operating status evaluation coefficient R(Yt,Hj,Fl), ensuring that its value range is (0,1); Use the fractional function to generate the performance status evaluation coefficient P(Jd, Sd, Fz, Nh, Sm), ensuring that its value range is (0, 1); Combine the operating status evaluation coefficient R (Yt, Hj, Fl) and the performance status evaluation coefficient P (Jd, Sd, Fz, Nh, Sm) to calculate the correlation screening coefficient Cρ between the two: Among them, L1 R and L2 P They are the positive adjustment coefficients of R(Yt,Hj,Fl) and P(Jd,Sd,Fz,Nh,Sm), 0.12≤L1 R ≤0.88, 0.11≤L2 p ≤0.92; Set the screening threshold of the correlation screening coefficient Cρ corresponding to the current time period T1 to C′ρ, where C′ρ is a positive number less than 1. When the correlation screening coefficient Cρ is greater than C′ρ, The accuracy, speed, load capacity, energy consumption and life indicators are taken as the key indicator set highly related to the equipment operation status data; among which η1, η2, η3, η4, η5 are the adjustment coefficients of the corresponding performance indicators, and the specific values ​​of η1, η2, η3, η4, η5 are obtained according to the test demonstration or the analysis of the expert group system; The performance indicators in the key indicator set are sorted from large to small according to the degree of correlation; the top three key performance indicators are represented by X1, X2, and X3 respectively; Step S3: Filter out the top three key performance indicators with the highest correlation with the equipment operation status data from the sorted key indicator set; Step S4: receiving and analyzing the key performance indicators obtained in step S3, generating an initial model, and optimizing the parameters of the initial model to obtain an industrial robot model; Step S5: In the subsequent time period T2 immediately adjacent to the current time period, the screening threshold of the current time period T1 is used as the preset threshold of the subsequent time period T2, and the ratio between the corresponding associated screening coefficients of the two adjacent time periods T1 and T2 is calculated. Based on the ratio, analysis is performed to determine whether it is necessary to optimize and adjust the preset threshold of the subsequent time period T2, so as to adaptively optimize the operating status parameters of the industrial robot model in different time periods.

2. The method for constructing an industrial robot model according to claim 1, characterized in that: Initial performance indicators include accuracy, speed, load capacity, energy consumption, and lifespan; Equipment operating status data includes operating time, environmental conditions, and load changes; Set the number of acquisitions in the current time period T1 to n, and obtain the acquisition number sequence {1, 2, ..., i, ..., n}, where i represents the index of the i-th acquisition number, and n represents the total number of acquisitions in the current time period; The generation of the correlation screening coefficient includes: Acquire equipment operation status data for analysis, and generate an operation status evaluation coefficient for graded evaluation of the operation status of the industrial robot; Acquire initial performance indicators for analysis, and generate performance status evaluation coefficients for grading and evaluating initial performance status of the industrial robot; Combine and analyze the operation status evaluation coefficient and the performance status evaluation coefficient to generate a correlation screening coefficient.

3. The method for constructing an industrial robot model according to claim 2, characterized in that: The explanations for the initial performance indicators and the equipment operation status data are as follows: Denote the precision as Jd. The precision Jd is quantitatively represented by taking the reciprocal of the operation deviation displacement between the actual operation of the robot and the preset target value through measurement. Denote the speed as Sd. The speed Sd is quantitatively represented by the average motion speed of the robot within the current time period T1. Denote the load capacity as Fz. The load capacity Fz is quantitatively represented by the maximum load mass that the robot can bear. Denote the energy consumption as Nh. The energy consumption Nh is quantitatively represented by the average energy consumption of the robot within the current time period T1. Denote the lifespan as Sm. The lifespan Sm is quantitatively represented by the expected service life of the robot under the preset working environment. Denote the operation time as Yt. The operation time Yt is quantitatively represented by the actual working time of the robot within the current time period T1. Denote the environmental conditions as Hj. The environmental conditions Hj include factors such as temperature, humidity, and the vibration amplitude of the installation part where the industrial robot is located. And Hj is comprehensively represented by the following calculation formula: Hj=w T ·Tue+Wed H ·SH+w V ·ZV Among them, WT, SH, and ZV represent the normalized values ​​of temperature, humidity, and vibration amplitude, respectively. T 、w H 、w V is the weight coefficient of the corresponding factor; Denote the load change as Fl. The load change Fl is quantitatively represented by the load change rate of the robot within the current time period T1.

4. The method for constructing an industrial robot model according to claim 3, characterized in that: The construction of the operation status evaluation coefficient specifically includes: Normalize the parameters included in the equipment operation status data collected within the current time period T1. Define the calculation formula of the operation status evaluation coefficient R(Yt, Hj, Fl) as follows: Among them, w t 、w e 、w l are the weight coefficients of operating time, environmental conditions, and load changes, respectively. The larger the output value of the operation status evaluation coefficient R(Yt, Hj, Fl), the better the operation status of the industrial robot.

5. The method for constructing an industrial robot model according to claim 4, characterized in that: Normalize the parameters included in the initial performance indicators collected within the current time period T1. Define the calculation formula of the performance status evaluation coefficient P(Jd, Sd, Fz, Nh, Sm) as follows: Among them, w Jd 、w Sd 、w Fz 、w Nh 、w Sm are the weight coefficients of accuracy, speed, load capacity, energy consumption, and life span, respectively. η8 is the correction term, η8 combined with the weight coefficient w Jd 、w Sd 、w Fz 、w Nh 、w Sm The adjustment is used to ensure the correct classification of the following four levels: high, medium-high, medium and low; according to the output value of the performance status evaluation coefficient P (Jd, Sd, Fz, Nh, Sm), the initial performance status of the industrial robot is evaluated at four levels: high, medium-high, medium and low; the specific classification method is as follows: When P(Jd, Sd, Fz, Nh, Sm) > 0.81, the initial performance status of the industrial robot is at a high level. When 0.65 < P(Jd, Sd, Fz, Nh, Sm) ≤ 0.81, the initial performance status of the industrial robot is at a medium-high level. When 0.36 < P(Jd, Sd, Fz, Nh, Sm) ≤ 0.65, the initial performance status of the industrial robot is at a medium level. When P(Jd, Sd, Fz, Nh, Sm) ≤ 0.36, the initial performance status of the industrial robot is at a low level.

6. The method for constructing an industrial robot model according to claim 5, characterized in that: Receive the generated initial model and conduct model construction to generate an industrial robot model, specifically including: Generate the initial model through the following formula: P total,n =a1×X1 n +a2×X2 n +a3×X3 n Among them, P total,n is the performance index value collected for the current nth time; a1, a2, a3 are the weight coefficients of the corresponding performance index, X1 n , X2 n 、X3 n They are the top three key performance indicators of correlation degree in the current nth collection; When P total,n The larger the value, the better the comprehensive performance of the industrial robot under the current conditions; When P total,n The smaller the value, the worse the overall performance of the industrial robot, and it needs to be optimized and adjusted; a1, a2, and a3 are all positive numbers, and a1 + a2 + a3 = 1; the values of a1, a2, and a3 are all within the range of 0 to 1. Optimize the weight coefficients of the initial model using an optimization algorithm. The optimization algorithm uses the gradient descent algorithm. And denote the updated and optimized values of the initialized weight coefficients a1, a2, and a3 as a1′, a2′, and a3′ in sequence. Apply the optimized weight coefficients a1′, a2′, and a3′ to the industrial robot model to form the final industrial robot model.

7. The method for constructing an industrial robot model according to claim 6, characterized in that: Calculate the ratio of the corresponding correlation screening coefficients of two adjacent time periods T1 and T2, and analyze based on the ratio to determine whether it is necessary to optimize and adjust the screening threshold of the next time period T2, including: The correlation screening coefficient Cρ calculation method of the current time period T1 is used to calculate and mark the correlation screening coefficient of the next time period T2 as Cρ2, and the ratio between Cρ and Cρ2 is calculated to obtain μ1 is the adjustment coefficient. The setting of μ1 is used to make The value range is limited to (0,1); and based on experimental demonstration or expert group analysis, the judgment threshold of the ratio is determined to be R1, R1∈(0,1); when When , the adjustment formula for C′ρ is: when When , the adjustment formula for C′ρ is: Wherein, C″ρ is the preset threshold after optimization adjustment.

8. A device for constructing an industrial robot model, characterized in that: The device is used to execute the method for constructing an industrial robot model according to any one of claims 1 to 7, comprising: Data acquisition module: used to collect the equipment operation status data and initial performance indicators of the industrial robot in the current time period T1; Key indicator set generation module: used to obtain the equipment operation status data and initial performance indicators of the current time period T1 for analysis, generate a correlation screening coefficient, and set a screening threshold of the correlation screening coefficient. The screening threshold is used to screen out a key indicator set that is highly correlated with the equipment operation status data from the initial performance indicators, and sort the various performance indicators in the key indicator set according to the degree of correlation; Screening module: used to screen out the top three key performance indicators with the highest correlation with the equipment operation status data from the sorted key indicator set; Initial model generation module: receives and analyzes the key performance indicators, generates an initial model, and optimizes the parameters of the initial model to obtain an industrial robot model; Optimization and adjustment module: It is used to use the screening threshold of the current time period T1 as the preset threshold of the next time period T2 in the next time period T2 immediately adjacent to the current time period, calculate the ratio between the corresponding associated screening coefficients of the two adjacent time periods T1 and T2, and analyze based on the ratio to determine whether it is necessary to optimize and adjust the screening threshold in the subsequent time period, so as to adaptively optimize the operating status parameters of the industrial robot model in different time periods.

9. A robot comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method for constructing an industrial robot model according to any one of claims 1 to 7 are implemented.

Images