A method for evaluating the comfort of operating equipment

Abstract

The present invention discloses a method for evaluating the comfort of an operating tool, which belongs to the field of power maintenance equipment optimization, and comprises the following steps: S1, collecting surface electromyographic signals of muscles of a tester when using an operating tool, and preprocessing the surface electromyographic signals to obtain a smooth signal set; S2, extracting and dividing the features of signals in the smooth signal set to obtain a training feature set and a test feature set; S3, constructing and training a fatigue recognition model according to the training feature set to obtain an optimal fatigue recognition model; S4, inputting the test feature set into the optimal fatigue recognition model to perform fatigue recognition to obtain fatigue time; S5, evaluating the comfort of the operating tool based on the fatigue time, and adjusting the weight and structure of the corresponding operating tool according to the comfort to obtain an optimal operating tool; the present invention selects the operating tool with the highest operating comfort by constructing the relationship between the environment, the tool and the physiological parameters of the operator, thereby significantly improving the work efficiency of the tester.

Description

Technical Field

[0001] The invention relates to the field of power maintenance equipment optimization, and in particular to a method for evaluating the comfort of operating equipment. Background Art

[0002] With the development of distribution network non-stop operation technology, and in order to avoid losses caused by power outage operations, the operation and maintenance workload of distribution network non-stop operations is increasing. However, at the current stage, when the distribution network is live, due to the high position of the distribution network conductors, weak bearing capacity of the conductors, and compact arrangement of electrical equipment, operation and maintenance personnel usually need to use operating tools when performing maintenance work. Because the environment in which the operation and maintenance personnel perform maintenance work is changeable and complex, it is difficult to quickly select operating tools that are highly comfortable and highly compatible with the operation and maintenance personnel's working posture according to the actual working environment, resulting in low operating efficiency of the operation and maintenance personnel and safety hazards.

[0003] Chinese patent, publication number: CN116196020A, publication date: June 2, 2023, discloses a patient muscle fatigue assessment system and method based on surface electromyography signals, including: a touch screen displays multiple limb movements for patients to select target limb movements, a multi-channel signal collector collects surface electromyography signals corresponding to each surface electromyography electrode within a set time at a set collection frequency, a multi-channel preprocessing module is used to preprocess the collected surface electromyography signals and transmit the preprocessed surface electromyography signals to a controller, the controller analyzes and corrects the preprocessed surface electromyography signals to obtain the patient's muscle fatigue level corresponding to this target limb movement, the controller controls the display screen to display the patient's muscle fatigue level corresponding to this target limb movement, and the controller controls the alarm to issue a corresponding warning when the patient's muscle fatigue level is above moderate fatigue; although the invention obtains the indirect reflection of the comfort of the human body's muscle fatigue for a certain action, it does not conduct in-depth analysis from the internal characteristics of the surface electromyography signal, resulting in the operating tools selected according to the muscle fatigue being unsuitable for the actual operation and maintenance environment. Summary of the invention

[0004] The purpose of the present invention is to address the problem that the prior art does not conduct in-depth analysis of the internal characteristics of surface electromyography signals, resulting in inappropriate selected operating tools; a method for evaluating the comfort of operating tools is proposed, which preprocesses the collected surface electromyography signals to eliminate interference signals in the collected signals, extracts features in the surface electromyography signals and inputs them into a fatigue recognition model for in-depth analysis to obtain the interaction force relationship between the weight and structure of the operating tool and the tester, and specifically quantifies the interaction force relationship using a comfort level. Finally, according to the comfort level, the weight and structure of the corresponding operating tool are adjusted to obtain the optimal operating tool that fits the tester's working posture best in the current working environment, thereby significantly improving the tester's working efficiency.

[0005] In a first aspect, a technical solution provided in an embodiment of the present invention is a method for evaluating the comfort of an operating device, comprising the following steps: S1, collecting surface electromyographic signals of muscles of the test personnel when using the operating tool in the current working environment, and preprocessing the surface electromyographic signals to obtain a smooth signal set; the surface electromyographic signals at least include bioelectric signals generated by muscles corresponding to muscle changes caused by the weight and structure of the operating tool; S2, extracting and dividing time domain features, frequency domain features and instantaneous frequency features based on the signals in the smooth signal set to obtain a training feature set and a test feature set; S3, constructing and training a fatigue recognition model according to the training feature set to obtain an optimal fatigue recognition model; S4, inputting the test feature set into the optimal fatigue recognition model to perform fatigue degree recognition to obtain fatigue time; S5. Evaluate the comfort of the operating tool based on the fatigue time, and adjust the weight and structure of the corresponding operating tool according to the comfort to obtain an optimal operating tool.

[0006] In this scheme, the collected surface electromyographic signals are preprocessed to eliminate interference signals in the collected signals to obtain a smooth signal set, and the smooth signal set is divided based on the time domain characteristics, frequency domain characteristics and instantaneous frequency characteristics of the signals in the smooth signal set to obtain a training feature set and a test feature set. A fatigue recognition model is constructed and trained based on the training feature set to obtain an optimal fatigue recognition model with the best performance suitable for the feature data. The test feature set is input into the optimal fatigue recognition model to perform deep fatigue recognition based on the internal characteristics of the feature data to obtain fatigue time, that is, when the tester uses the operating tool in the current working environment, the surface electromyographic signal of the muscle corresponding to the muscle change shows the time when the tester enters the fatigue state. At the same time, the comfort of the operating tool is evaluated based on the fatigue time, and the interaction force relationship between operating tools of different weights and structures and the tester is quantified. Finally, the weight and structure of the corresponding operating tool are adjusted according to the comfort to obtain the optimal operating tool with the highest fit to the tester's working posture in the current working environment, which significantly improves the tester's working efficiency and effectively avoids safety hazards caused by inappropriate operating tools.

[0007] Preferably, in S1, the specific process of preprocessing the surface electromyography signal to obtain a smooth signal set is: S111, filtering the surface electromyographic signal with a bandpass filter, and performing reverse smoothing processing on the filtered surface electromyographic signal to obtain a smoothed electrical signal; S112, performing empirical mode decomposition on the smoothed electrical signal to obtain intrinsic mode function components; S113, sorting the smoothed electrical signal and the intrinsic mode function components to obtain a smoothed signal set.

[0008] In this scheme, after the filtered surface electromyographic signal is subjected to reverse smoothing and empirical mode decomposition, the internal characteristics of the surface electromyographic signal are preliminarily analyzed and exposed, so as to stabilize the surface electromyographic signal and effectively improve the accuracy of the surface electromyographic signal characteristics extracted subsequently.

[0009] Preferably, in S111, the formula for reverse smoothing is: Wherein, d(n) represents the surface electromyographic signal after reverse difference operation, ABS represents the absolute value function, k is an integer, and the minimum value of k is 1 and the maximum value is 4, y(n) represents the surface electromyographic signal after the nth sampling point is sampled and passed through the bandpass filter, x(n) represents the smoothed electrical signal, and N is 10.

[0010] Preferably, in S112, the smoothed electrical signal is subjected to empirical mode decomposition to obtain intrinsic mode function components, and the intrinsic mode function components need to be screened using correlation coefficients, and the calculation formula of the correlation coefficients is specifically: Where L is the length of the surface electromyography signal, r i is the correlation coefficient between the ith intrinsic mode function component and the surface electromyographic signal x(j), IMF i (j) is the i-th intrinsic mode function component corresponding to the surface electromyography signal x(j).

[0011] In this scheme, by setting a corresponding threshold value as a screening condition based on the correlation coefficient, the pseudo components generated by the smoothed electrical signal during the empirical mode decomposition process can be eliminated, effectively improving the accuracy and effectiveness of the intrinsic mode function components obtained by performing empirical mode decomposition on the smoothed electrical signal.

[0012] Preferably, in S2, the time domain feature extraction formula is specifically: Where RMS is the root mean square of the smoothed electrical signal in the smoothed signal set, x i is the sampling value of the surface electromyography signal at the i-th sampling point, and N is the length of the electromyography signal segment of the surface electromyography signal.

[0013] Preferably, in S2, the frequency domain feature extraction formula is specifically: Where f1 is the first cutoff frequency of the hardware, f2 is the second cutoff frequency of the hardware, MDF is the median frequency of the frequency domain feature, and P(f) is the power spectral density obtained by performing spectrum transformation on the smoothed electrical signal in the smoothed signal set.

[0014] Preferably, in S2, the process of extracting the instantaneous frequency feature is specifically as follows: S231, performing Hilbert transform on the intrinsic mode function components in the smooth signal set to obtain instantaneous phase and instantaneous amplitude; S232, obtaining instantaneous frequency based on the derivative of the instantaneous phase, and weighting the instantaneous frequency with the instantaneous amplitude as the weight to obtain the average instantaneous frequency.

[0015] Preferably, in S3, the fatigue recognition model constructed according to the training feature set is specifically: In the formula, max represents the objective function of the fatigue recognition model, m is the interval of the feature space, and a i is a Lagrangian operator greater than zero, a j is another Lagrangian operator, K(.) is the kernel function of the fatigue identification model, x i is the i-th sample, y i For sample x i The category label, x j is the jth sample, y j For sample x j , st represents the constraint condition of the fatigue recognition model, and c is the error penalty parameter.

[0016] In this scheme, the fatigue recognition model analyzes the surface electromyographic signals of different test persons for operating instruments of the same weight and structure according to the characteristics of electromyographic signals, and obtains the fatigue time of test persons with different physical fitness under the same operating instrument conditions and working environment. The fatigue recognition model also analyzes the surface electromyographic signals of the same test person for operating instruments of different weights and structures, and obtains the fatigue time of test persons with the same physical fitness under different operating instrument conditions and working environments, thereby significantly improving the reliability of the fatigue time.

[0017] Preferably, in S3, the specific process of training the fatigue recognition model according to the training feature set to obtain the optimal fatigue recognition model is: S321, initializing a population of a fatigue recognition model based on the training feature set, and calculating the fitness of individuals in the population; S322, dividing the population into a first subpopulation and a second subpopulation based on the fitness of the individuals in the population; S323, obtaining a first optimal solution according to the first subpopulation, and obtaining a second optimal solution according to the second subpopulation; S324. Integrate the first subpopulation and the second subpopulation to obtain a new population based on the first optimal solution and the second optimal solution, and recalculate the fitness of the individuals according to the integrated new population to obtain an optimal penalty parameter and an optimal kernel function; S325. Adjust the fatigue recognition model according to the optimal penalty parameter and the optimal kernel function to obtain an optimal fatigue recognition model.

[0018] In this solution, a fatigue recognition model is trained based on the training feature set, and the fatigue recognition model is adjusted according to the optimal penalty parameter and the optimal kernel function found during the training process to obtain an optimal fatigue recognition model with the best model performance, thereby effectively improving the reliability of the optimal fatigue recognition model.

[0019] Preferably, in S5, the comfort level of the operating tool is evaluated based on the fatigue time by normalization processing, and the calculation formula of the normalization processing is specifically: In the formula, c ij is the comfort index of test person i when using operating device j, t ij is the time when tester i enters fatigue state when using operating device j, min(t i ) is the minimum time for tester i to enter fatigue state using the operating device, max(t i ) is the maximum time that the test person i enters the fatigue state when using the operating equipment.

[0020] In this embodiment, the comfort level of the operating device is evaluated based on a calculation formula based on normalization processing combined with fatigue time, which significantly reduces the evaluation error caused by different physical fitness of testers.

[0021] The present invention has at least the following substantial effects: This scheme preprocesses the collected surface electromyography signals to eliminate interference signals in the collected signals to obtain a smooth signal set, and uses the correlation coefficient to set a threshold to eliminate pseudo components that appear in the preprocessing process to ensure the reliability of the smooth data set; when the smooth data set is constructed, the time domain characteristics, frequency domain characteristics and instantaneous frequency characteristics of the signals in the smooth signal set are extracted to obtain a training feature set and a test feature set, and a fatigue recognition model is constructed and trained based on the training feature set to obtain an optimal fatigue recognition model with the best performance suitable for the feature data, and the test feature set is input into the optimal fatigue recognition model to perform deep fatigue recognition based on the internal characteristics of the feature data to obtain the fatigue time That is, when the tester uses the operating tool in the current working environment, the surface electromyographic signal of the muscle corresponding to the muscle change is used to display the time when the tester enters the fatigue state. At the same time, the comfort of the operating tool is evaluated according to the fatigue time and the normalized processing calculation formula, and the interaction force relationship between the operating tools of different weights and structures and the tester is quantified, and the error caused by the physical fitness of the tester is reduced. Finally, the weight and structure of the corresponding operating tool are adjusted according to the comfort level to obtain the optimal operating tool with the highest fit to the tester's working posture in the current working environment, which significantly improves the tester's working efficiency and effectively avoids safety hazards caused by inappropriate operating tools. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Other features, objects and advantages of the present invention will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings. The drawings are only for the purpose of illustrating preferred embodiments and are not to be considered as limiting the present invention. Also, the same reference symbols are used throughout the drawings to represent the same parts.

[0023] Figure 1 This is a flow chart of a method for evaluating the comfort level of an operating device according to the present invention; Figure 2 Schematic diagram of obtaining the intrinsic mode function components by performing empirical mode decomposition for smoothing electrical signals. DETAILED DESCRIPTION

[0024] 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 the accompanying drawings and embodiments. It should be understood that the specific implementation method described herein is only an optimal embodiment of the present invention, which is only used to explain the present invention and does not limit the scope of protection of the present invention. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0025] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the operations (or steps) as sequential processes, many of the operations (or steps) therein can be implemented in parallel, concurrently, or simultaneously. In addition, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but can also have additional steps not included in the drawings; the process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0026] Embodiment 1: like Figure 1 As shown, this embodiment provides a method for evaluating the comfort of operating an appliance, comprising the following steps: S1. Collect the surface electromyographic signals of the muscles of the test personnel when using the operating tool in the current working environment, and pre-process the surface electromyographic signals to obtain a smooth signal set; the surface electromyographic signals at least include bioelectric signals generated by the muscles corresponding to the muscle changes caused by the weight and structure of the operating tool.

[0027] The specific process of preprocessing the surface electromyography signal to obtain a smooth signal set is as follows: S111, using a bandpass filter to filter the surface electromyographic signal, and performing reverse smoothing processing on the filtered surface electromyographic signal to obtain a smoothed electrical signal.

[0028] In S111, the formula for reverse smoothing is specifically: Wherein, d(n) represents the surface electromyographic signal after reverse difference operation, ABS represents the absolute value function, k is an integer, and the minimum value of k is 1 and the maximum value is 4, y(n) represents the surface electromyographic signal after the nth sampling point is sampled and passed through the bandpass filter, x(n) represents the smoothed electrical signal, and N is 10.

[0029] S112, performing empirical mode decomposition on the smoothed electrical signal to obtain intrinsic mode function components.

[0030] In S112, the smoothed electrical signal is subjected to empirical mode decomposition to obtain intrinsic mode function components, and the intrinsic mode function components need to be screened using correlation coefficients. The calculation formula of the correlation coefficient is specifically: Where L is the length of the surface electromyography signal, r i is the correlation coefficient between the ith intrinsic mode function component and the surface electromyographic signal x(j), IMF i(j) is the i-th intrinsic mode function component corresponding to the surface electromyography signal x(j).

[0031] S113, sorting the smoothed electrical signal and the intrinsic mode function components to obtain a smoothed signal set.

[0032] In this embodiment, the amplitude of the surface electromyographic signal is generally 0-5mV, and the frequency is mainly concentrated in 5-500Hz. It is a weak bioelectric signal generated when biological muscles change. It is easily affected by factors such as moving artifact noise, environmental noise, and inherent noise of instrument electronic components during collection, resulting in interference electrical signals. Therefore, a 10-500Hz bandpass filter is used for filtering in this scheme. However, a 50Hz notch filter for filtering the power frequency will filter out some key information of the surface electromyographic signal and cannot be used. After the filtering is completed, the reverse smoothing processing formula in S111 is required to perform reverse difference operation and smoothing filtering on the surface electromyographic signal to further remove high-frequency noise and power frequency interference. At this point, a surface electromyographic signal that is not affected by interference factors, that is, a smoothed electrical signal, can be obtained. However, the surface electromyographic signal is a nonlinear and non-stationary electrical signal. Directly processing and extracting features from it is likely to cause errors, resulting in failure of subsequent analysis. Therefore, it is necessary to use the empirical mode decomposition method to decompose the smoothed electrical signal into the following: Figure 2 The intrinsic mode function components shown in the figure can be expressed as follows after decomposition and smoothing: In the formula, x(t) represents the smoothed electrical signal corresponding to time t, c i (t) represents the i-th eigenmode function component, r n (t) represents the residual function, n represents the number of decompositions of the intrinsic mode function components; The intrinsic mode function components during the preliminary decomposition also need to meet the following requirements: ① the number of extreme values ​​and the number of zero crossings must be equal to or differ by 1; ② the average value of the local minimum value and the upper and lower envelope limits defined by the minimum value must be zero; when the above two conditions are met, part of the intrinsic mode function components at this time are pseudo components generated during the decomposition of the smoothed electrical signal. A threshold value can be set to filter the intrinsic mode function components to eliminate pseudo components. The threshold value is set to one tenth of the maximum value in the correlation coefficient sequence.

[0033] Taking the surface electromyographic signal corresponding to the biceps brachii of the right hand when the tester 1 uses the ratchet hanger to replace the tension insulator as an example, the correlation coefficients between each IMF component, that is, the intrinsic mode function component (excluding the residual function) and the surface electromyographic signal are shown in Table 1: Table 1. Correlation coefficients between IMF components and surface electromyographic signals IMF Components IMF1 IMF2 IMF3 IMF4 IMF5 IMF6 IMF7 IMF8 Correlation coefficient 0.8629 0.3452 0.1419 0.0252 0.0056 0.0068 0.0034 0.0024 The threshold at this time is 0.086.

[0034] S2. Extract and divide time domain features, frequency domain features and instantaneous frequency features based on the signals in the smooth signal set to obtain a training feature set and a test feature set.

[0035] The specific formula for extracting the time domain feature is: Where RMS is the root mean square of the smoothed electrical signal in the smoothed signal set, x i is the sampling value of the surface electromyography signal at the i-th sampling point, and N is the length of the electromyography signal segment of the surface electromyography signal.

[0036] The frequency domain feature extraction formula is specifically: Where f1 is the first cutoff frequency of the hardware, f2 is the second cutoff frequency of the hardware, MDF is the median frequency of the frequency domain feature, and P(f) is the power spectral density obtained by performing spectrum transformation on the smoothed electrical signal in the smoothed signal set.

[0037] The extraction process of the instantaneous frequency feature is specifically as follows: S231, performing Hilbert transform on the intrinsic mode function components in the smooth signal set to obtain instantaneous phase and instantaneous amplitude; S232, obtaining instantaneous frequency based on the derivative of the instantaneous phase, and weighting the instantaneous frequency with the instantaneous amplitude as the weight to obtain the average instantaneous frequency.

[0038] In this embodiment, the Hilbert transform is introduced to more accurately extract the average instantaneous frequency AIF of the surface electromyographic signal. First, the IMF component c i (t) respectively perform Hilbert transform, which is defined as: Where, d i (t) is the transformation corresponding to the i-th IMF component at time t, c i (τ) is the i-th IMF component at time τ; Based on the above definition, the analytical signal corresponding to the IMF component can be obtained as: In the formula, Z i (t) is the analytical signal corresponding to the i-th IMF component at time t, c i (t) is the i-th IMF component at time t, d i (t) is the transformation corresponding to the i-th IMF component at time t, a i (t) is the instantaneous amplitude, θ i (t) is the instantaneous phase; the calculation formula of the instantaneous amplitude and the instantaneous phase is: Then the instantaneous phase θ i (t) by taking the derivative, we can get the instantaneous frequency of the IMF component, the instantaneous frequency ω i (t) is as follows: Finally, with the instantaneous amplitude a i (t) is the weight, and the average instantaneous frequency AIF of each IMF component at each m point is calculated respectively i , the average instantaneous frequency AIF i As shown below: In the formula, ω i (j) is the instantaneous frequency of point j, is the square of the instantaneous amplitude at point j; Then, the average instantaneous frequency AIF of every m points can be obtained, and the average instantaneous frequency AIF is as follows: Where AIF i is the average instantaneous frequency of each IMF component at every m points, a i is the instantaneous amplitude.

[0039] S3. Constructing and training a fatigue recognition model according to the training feature set to obtain an optimal fatigue recognition model.

[0040] The fatigue recognition model constructed according to the training feature set is specifically: In the formula, max represents the objective function of the fatigue recognition model, m is the interval of the feature space, and a i is a Lagrangian operator greater than zero, a j is another Lagrangian operator, K(.) is the kernel function of the fatigue identification model, x i is the i-th sample, y i For sample x i The category label, x j is the jth sample, y j For sample x j , st represents the constraint condition of the fatigue recognition model, and c is the error penalty parameter.

[0041] In this embodiment, the constructed fatigue recognition model is a supervised learning model SVM with a binary classification design. Compared with other classification models, SVM can create a hyperplane in a high-dimensional space to classify data that cannot be linearly separated. Combined with the different features of the surface electromyography signal extracted above, these features do not have the basic conditions for linear separation from each other, so the supervised learning model SVM is selected as the fatigue recognition model. Secondly, the error penalty parameter c in the model can control the degree of penalty for wrongly classified samples. The division ratio of the training feature set and the test feature set can be adjusted according to actual needs. In this embodiment, the training feature set accounts for 70% and the test feature set accounts for 30%.

[0042] The specific process of training the fatigue recognition model according to the training feature set to obtain the optimal fatigue recognition model is as follows: S321, initializing a population of a fatigue recognition model based on the training feature set, and calculating the fitness of individuals in the population; S322, dividing the population into a first subpopulation and a second subpopulation based on the fitness of the individuals in the population; S323, obtaining a first optimal solution according to the first subpopulation, and obtaining a second optimal solution according to the second subpopulation; S324. Integrate the first subpopulation and the second subpopulation to obtain a new population based on the first optimal solution and the second optimal solution, and recalculate the fitness of the individuals according to the integrated new population to obtain an optimal penalty parameter and an optimal kernel function; S325. Adjust the fatigue recognition model according to the optimal penalty parameter and the optimal kernel function to obtain an optimal fatigue recognition model.

[0043] In this embodiment, a PSO-GA hybrid optimization method based on traditional metaheuristic algorithms PSO and GA is selected to train a fatigue recognition model. By comparing the PSO-GA hybrid optimization method with traditional metaheuristic algorithms PSO and GA, it can be seen that: ① The local optimization ability of the traditional metaheuristic algorithm GA is outstanding, but the calculation is time-consuming and sensitive to the initial population, which easily leads to local optimal solutions; ② The traditional metaheuristic algorithm PSO has better global optimization ability and fast calculation speed, but its local optimization ability is poor and it is prone to premature convergence; The PSO-GA hybrid optimization method combines the advantages of the two, and has both excellent global optimization ability and calculation speed, as well as excellent local optimization ability; secondly, the fitness is the average accuracy of the fatigue recognition model calculated by the ten-fold cross validation method.

[0044] S4. Input the test feature set into the optimal fatigue recognition model to perform fatigue degree recognition to obtain fatigue time.

[0045] S5. Evaluate the comfort of the operating tool based on the fatigue time, and adjust the weight and structure of the corresponding operating tool according to the comfort to obtain an optimal operating tool.

[0046] The fatigue time is used to evaluate the comfort of the operating device by normalization processing, and the calculation formula of the normalization processing is specifically: In the formula, c ij is the comfort index of test person i when using operating device j, t ij is the time when tester i enters fatigue state when using operating device j, min(t i ) is the minimum time for tester i to enter fatigue state using the operating device, max(t i ) is the maximum time that the test person i enters the fatigue state when using the operating equipment.

[0047] In this embodiment, the comfort index is defined by the time it takes for the corresponding muscles of the tester to enter a fatigue state when using the operating tool. Due to the different physical fitness of the testers, even if the objective conditions are completely consistent, there will be large differences in the time scale of the fatigue state of different testers. Therefore, a normalized processing method is used to estimate the comfort.

[0048] In order to verify the actual effect of this embodiment, this embodiment evaluates the comfort of testers using insulating rods of different weights and compares and analyzes the actual feelings of the testers; The comfort evaluation data of the insulating rods of different weights are shown in Table 2: Table 2. Comfort level of testers using different weights of insulating rods to move up and down Weight / kg 0.58 0.72 0.83 0.92 1.15 1.27 1.35 Biceps comfort 1.00 0.951 0.929 0.891 0.753 0.649 0.512 Triceps comfort 1.00 1.00 0.949 0.921 0.851 0.728 0.591 Deltoid comfort 1.00 1.00 0.963 0.945 0.856 0.742 0.642 Brachioradialis comfort 1.00 0.958 0.924 0.902 0.769 0.693 0.567 Average comfort 1.00 0.977 0.941 0.915 0.807 0.703 0.578 Minimum comfort 1.00 0.951 0.924 0.891 0.753 0.649 0.512 Maximum comfort 1.00 1.00 0.963 0.945 0.856 0.742 0.642 It can be seen from the data in the table that with the increase of the weight of the insulating rod, the comfort of each muscle of the tester generally decreases, and the comfort of each muscle is: deltoid > triceps > brachioradialis > biceps, among which the comfort of biceps and brachioradialis is lower than the average comfort, and the comfort of deltoid and triceps is higher than the average comfort; in the process of up and down reciprocating motion, elbow flexion and extension actions are mainly involved, including main muscles such as brachialis, biceps, brachioradialis, triceps and anconeus. In the specific elbow flexion and extension actions, the tester mainly controls the insulating rod through the brachialis, biceps and brachioradialis to complete the elbow flexion action. Among them, the biceps and brachioradialis, as the main muscles of elbow flexion, bear a greater load and are more likely to enter a fatigue state, while the elbow extension action does not simulate the entire elbow extension process, and the triceps does not reach the maximum elbow extension, so it is difficult to enter a fatigue state.

[0049] Comparing the average comfort of insulating rods of different weights, it can be seen that when the weight increases from 0.58kg to 0.92kg, the amplitude of each decrease in the average comfort is about 0.03, and when the weight increases from 0.92kg to 1.35kg, the amplitude of each decrease in the average comfort is above 0.1; among them, when the weight increases from 0.92kg to 1.15kg, the average comfort decreases from 0.915 to 0.807, and the decrease reaches 0.108. At this time, the muscles of the test personnel have reached the fatigue limit, causing the muscles of the test personnel to quickly enter the fatigue state, and the decrease in comfort becomes larger; in addition, when the weight of the insulating rod is 0.58kg, its average comfort is the largest, and none of the muscles enters the fatigue state. When the weight of the insulating rod is 0.72kg, the triceps and deltoid muscles do not enter the fatigue state, and the brachioradialis and biceps enter the fatigue state earlier than when it is 0.58kg, and the comfort decreases by 5%.

[0050] In summary, when the testers used an insulating rod weighing less than 0.72 kg, the comfort level of each muscle was within an acceptable range. Therefore, different materials can be selected to control the weight of the insulating rod within 0.72 kg.

[0051] This embodiment has at least the following substantial effects: In this embodiment, the collected surface electromyography signals are preprocessed to eliminate interference signals in the collected signals to obtain a smooth signal set, and the correlation coefficient is used to set a threshold to eliminate pseudo components that appear in the preprocessing process to ensure the reliability of the smooth data set; when the smooth data set is constructed, the time domain characteristics, frequency domain characteristics and instantaneous frequency characteristics of the signals in the smooth signal set are extracted to obtain a training feature set and a test feature set, and a fatigue recognition model is constructed and trained based on the training feature set to obtain an optimal fatigue recognition model with the best performance suitable for the feature data, and the test feature set is input into the optimal fatigue recognition model to perform deep fatigue recognition based on the internal characteristics of the feature data to obtain the fatigue time. That is, when the tester uses the operating tool in the current working environment, the surface electromyographic signal of the muscle corresponding to the muscle change is used to display the time when the tester enters the fatigue state. At the same time, the comfort of the operating tool is evaluated according to the fatigue time and the normalized processing calculation formula, and the interaction force relationship between the operating tools of different weights and structures and the tester is quantified, and the error caused by the physical fitness of the tester is reduced. Finally, the weight and structure of the corresponding operating tool are adjusted according to the comfort level to obtain the optimal operating tool with the highest fit to the tester's working posture in the current working environment, which significantly improves the tester's working efficiency and effectively avoids safety hazards caused by inappropriate operating tools.

[0052] The above specific embodiments are preferred embodiments of the present invention, and are not intended to limit the specific implementation scope of the present invention. The scope of the present invention includes but is not limited to the present specific embodiments. All equivalent changes made in accordance with the shape, structure, and method of the present invention are within the protection scope of the present invention.

Claims

1. A method for evaluating the comfort of an operating device, characterized in that: The method comprises the following steps: S1, collecting surface electromyographic signals of muscles of the test personnel when using the operating tool in the current working environment, and preprocessing the surface electromyographic signals to obtain a smooth signal set; the surface electromyographic signals at least include bioelectric signals generated by muscles corresponding to muscle changes caused by the weight and structure of the operating tool; S2, extracting and dividing time domain features, frequency domain features and instantaneous frequency features based on the signals in the smooth signal set to obtain a training feature set and a test feature set; S3, constructing and training a fatigue recognition model according to the training feature set to obtain an optimal fatigue recognition model; S4, inputting the test feature set into the optimal fatigue recognition model to perform fatigue degree recognition to obtain fatigue time; S5. Evaluate the comfort level of the operating tool based on the fatigue time, and adjust the weight and structure of the corresponding operating tool according to the comfort level to obtain an optimal operating tool.

2. The method for evaluating the comfort of an operating device according to claim 1, characterized in that: In S1, the specific process of preprocessing the surface electromyography signal to obtain a smooth signal set is: S111, filtering the surface electromyographic signal with a bandpass filter, and performing reverse smoothing processing on the filtered surface electromyographic signal to obtain a smoothed electrical signal; S112, performing empirical mode decomposition on the smoothed electrical signal to obtain intrinsic mode function components; S113, sorting the smoothed electrical signal and the intrinsic mode function components to obtain a smoothed signal set.

3. The method for evaluating the comfort of an operating device according to claim 2, characterized in that: In S111, the formula for reverse smoothing is specifically: Wherein, d(n) represents the surface electromyographic signal after reverse difference operation, ABS represents the absolute value function, k is an integer, and the minimum value of k is 1 and the maximum value is 4, y(n) represents the surface electromyographic signal after the nth sampling point is sampled and passed through the bandpass filter, x(n) represents the smoothed electrical signal, and N is 10.

4. The method for evaluating the comfort of an operating device according to claim 2, characterized in that: In S112, the smoothed electrical signal is subjected to empirical mode decomposition to obtain intrinsic mode function components, and the intrinsic mode function components need to be screened using correlation coefficients. The calculation formula of the correlation coefficient is specifically: Where L is the length of the surface electromyography signal, r i is the correlation coefficient between the ith intrinsic mode function component and the surface electromyographic signal x(j), IMF i (j) is the i-th intrinsic mode function component corresponding to the surface electromyography signal x(j).

5. The method for evaluating the comfort of an operating device according to claim 1, characterized in that: In S2, the time domain feature extraction formula is specifically: Where RMS is the root mean square of the smoothed electrical signal in the smoothed signal set, x i is the sampling value of the surface electromyographic signal at the i-th sampling point, and N is the length of the electromyographic signal segment of the surface electromyographic signal.

6. The method for evaluating the comfort of an operating device according to claim 1, characterized in that: In S2, the frequency domain feature extraction formula is specifically: Where f1 is the first cutoff frequency of the hardware, f2 is the second cutoff frequency of the hardware, MDF is the median frequency of the frequency domain feature, and P(f) is the power spectral density obtained by performing spectrum transformation on the smoothed electrical signal in the smoothed signal set.

7. The method for evaluating the comfort of an operating device according to claim 1, characterized in that: In S2, the process of extracting the instantaneous frequency feature is specifically as follows: S231, performing Hilbert transform on the intrinsic mode function components in the smooth signal set to obtain instantaneous phase and instantaneous amplitude; S232: Obtain an instantaneous frequency based on the instantaneous phase derivative, and perform weighted calculation on the instantaneous frequency using the instantaneous amplitude as a weight to obtain an average instantaneous frequency.

8. The method for evaluating the comfort of an operating device according to claim 1, characterized in that: In S3, the fatigue recognition model constructed according to the training feature set is specifically: In the formula, max represents the objective function of the fatigue recognition model, m is the interval of the feature space, and a i is a Lagrangian operator greater than zero, a j is another Lagrangian operator, K(.) is the kernel function of the fatigue identification model, x i is the i-th sample, y i For sample x i The category label of x j is the jth sample, y j For sample x j , st represents the constraint condition of the fatigue recognition model, and c is the error penalty parameter.

9. The method for evaluating the comfort of an operating device according to claim 1, characterized in that: In S3, the specific process of training the fatigue recognition model according to the training feature set to obtain the optimal fatigue recognition model is: S321, initializing a population of a fatigue recognition model based on the training feature set, and calculating the fitness of individuals in the population; S322, dividing the population into a first subpopulation and a second subpopulation based on the fitness of the individuals in the population; S323, obtaining a first optimal solution according to the first subpopulation, and obtaining a second optimal solution according to the second subpopulation; S324. Integrate the first subpopulation and the second subpopulation to obtain a new population based on the first optimal solution and the second optimal solution, and recalculate the fitness of the individuals according to the integrated new population to obtain an optimal penalty parameter and an optimal kernel function; S325. Adjust the fatigue recognition model according to the optimal penalty parameter and the optimal kernel function to obtain an optimal fatigue recognition model.

10. The method for evaluating the comfort of an operating device according to claim 1, characterized in that: In S5, the comfort level of the operating tool is evaluated based on the fatigue time by normalization processing, and the calculation formula of the normalization processing is specifically: In the formula, c ij is the comfort index of test person i when using operating device j, t ij is the time when tester i enters fatigue state when using operating device j, min(t i ) is the minimum time for tester i to enter fatigue state using the operating device, max(t i ) is the maximum time that the test person i enters the fatigue state when using the operating equipment.

Images