Method and device for identifying working condition of excavator

Abstract

The present application discloses a method for identifying the working condition of an excavator and an apparatus for identifying the working condition of an excavator. The method comprises: obtaining the pilot control signal data and dual pump pressure signal data of the operating handle of the excavator; preprocessing the pilot control signal data and dual pump pressure signal data of the operating handle to obtain preprocessed data; constructing a feature vector based on the preprocessed data; performing feature dimensionality reduction on the feature vector to obtain a reduced dimensionality feature vector; inputting the reduced dimensionality feature vector into a classification model to obtain a first working condition mode of the excavator; verifying the first working condition mode and outputting a detailed list of working condition identification results. The present application constructs a feature vector based on the basic time domain characteristics of the composite signal and the actual mining situation, and performs dimensionality reduction on the feature vector to remove feature factors that have little influence on the identification result. After the model identification is completed, the classification result is verified to correct the model identification result and improve the classification accuracy of the model.

Description

Technical Field

[0001] The present application relates to the technical field of excavator pattern recognition, and in particular to a method for identifying an excavator operating condition and an excavator operating condition identification device. Background Art

[0002] Excavators are essential construction machinery for infrastructure development and are widely used in mining, construction, water conservancy projects, and other fields. Currently, excavators suffer from low operating efficiency and irrational task planning, which seriously impact their lifespan. To better demonstrate the versatility and efficient operation of excavators, it is necessary to identify each excavator's operating mode and monitor its fuel consumption to achieve energy conservation, control, and emission reduction goals. Furthermore, more rational construction task planning can be implemented to fully utilize the excavator's operational capabilities.

[0003] Currently, there are two main methods for identifying the working conditions of hydraulic excavators. The first method is mainly image recognition. This method requires the installation of image or video acquisition equipment at the construction site and the use of image processing algorithms such as target detection to identify and measure the working conditions of the excavator. This method is relatively expensive. The second method is to construct a classification model based on the main pump pressure and pilot signal for identification. However, the current modeling basically extracts the statistical characteristic values ​​of the samples, such as the mean and variance, which are widely used features. However, these cannot well reflect the excavation and leveling modes, and the model accuracy is low. At the same time, when using machine learning algorithms, there are few methods to explain how to select the algorithm parameters, which can easily lead to problems such as low model learning ability or overfitting. Summary of the Invention

[0004] The purpose of the embodiments of the present application is to provide a method and an apparatus for identifying the working condition of an excavator, so as to solve the problem of low accuracy in identifying the working condition pattern of an excavator in the prior art.

[0005] To achieve the above-mentioned objectives, the present application provides, in a first aspect, a method for identifying an excavator working condition, which is applied to an excavator working condition identification device. The method comprises:

[0006] Obtain pilot control signal data of the excavator's operating handle and dual pump pressure signal data;

[0007] Preprocessing the operating handle pilot control signal data and the dual pump pressure signal data to obtain preprocessed data;

[0008] Construct feature vectors based on preprocessed data;

[0009] Performing feature dimensionality reduction on the feature vector to obtain a reduced-dimensional feature vector;

[0010] Inputting the reduced-dimensional feature vector into a classification model to obtain a first working mode of the excavator;

[0011] Verify the first operating mode and output a detailed list of operating mode identification results.

[0012] In an embodiment of the present application, the method further includes:

[0013] Obtain the excavator's crushing working mode status, machine construction status, machine travel status and engine speed data;

[0014] The second working mode of the excavator is determined according to the crushing working mode status, machine construction status, machine walking status and engine speed data, and a detailed list of working mode identification results is output.

[0015] In an embodiment of the present application, the classification model is a libsvm model, and the method further includes:

[0016] Determine the target parameters of the libsvm model to build a classification model.

[0017] In an embodiment of the present application, determining target parameters of the libsvm model to construct a classification model includes:

[0018] Encode the initial parameters of the libsvm model to form an initial population, which includes multiple individuals;

[0019] Determine the fitness of each individual in the initial population separately;

[0020] Determine whether there is an individual whose fitness meets the termination conditions;

[0021] When it is determined that there is an individual whose fitness satisfies the termination condition, the parameters corresponding to the individual whose fitness satisfies the termination condition are determined as the target parameters of the libsvm model;

[0022] When it is determined that there is no individual whose fitness satisfies the termination condition, the initial population is updated by performing selection, crossover and mutation through genetic operators until there is an individual whose fitness satisfies the termination condition.

[0023] In the embodiment of the present application, preprocessing the operating handle pilot control signal data and the dual pump pressure signal data to obtain preprocessed data includes:

[0024] Performing data filling processing on the pilot control signal data of the operating handle and the dual pump pressure signal data to obtain filled data;

[0025] Performing data alignment processing on the padded data to obtain aligned data;

[0026] The aligned data are subjected to data denoising to obtain preprocessed data.

[0027] In the embodiment of the present application, the feature vector includes a first feature vector and a second feature vector, and constructing the feature vector based on the preprocessed data includes:

[0028] Get the pressure band of the excavator;

[0029] Intercept the pressure band of the preset size window;

[0030] Extracting the time domain features of the pressure band of a preset size window to obtain a first feature vector;

[0031] The pressure ratios of the excavation signal, the unloading signal, and the rotation signal in the pressure band of the preset size window are extracted to obtain a second eigenvector.

[0032] In the embodiment of the present application, performing feature dimensionality reduction on the feature vector to obtain a reduced-dimensional feature vector includes:

[0033] Input the eigenvector and the operating mode label into the random forest function to obtain the contribution rate of each eigenvector;

[0034] Sort the feature vectors according to contribution rate and preset order;

[0035] The preset number of feature vectors is determined as the feature vectors after dimensionality reduction.

[0036] In the embodiment of the present application, verifying the first operating mode includes:

[0037] Obtain the peak and trough vibration frequencies of the excavation pilot signal, unloading pilot signal, rotation pilot signal, arm extension pilot signal, and arm retraction pilot signal of the excavator;

[0038] The working condition identification result is verified based on the peak and trough vibration frequencies of the excavation pilot signal, the unloading pilot signal, the rotation pilot signal, the bucket arm extension pilot signal and the bucket arm retraction pilot signal.

[0039] A second aspect of the present application provides an excavator working condition identification device, comprising:

[0040] a memory configured to store instructions; and

[0041] The processor is configured to call the instructions from the memory and implement the above-mentioned method for identifying the working condition of the excavator when executing the instructions.

[0042] A third aspect of the present application provides a machine-readable storage medium having stored thereon instructions for causing a machine to execute the above-mentioned method for identifying the working condition of an excavator.

[0043] Through the above technical solution, the pilot control signal data and dual pump pressure signal data of the operating handle of the excavator are first obtained, and the pilot control signal data and dual pump pressure signal data of the operating handle are preprocessed to obtain preprocessed data. Then, a feature vector is constructed based on the preprocessed data, and the feature vector is subjected to feature dimensionality reduction to obtain a reduced feature vector. Finally, the reduced feature vector is input into the classification model to obtain the first working mode of the excavator, and the first working mode is verified and a detailed table of working condition identification results is output. This application constructs a feature vector based on the basic time domain characteristics of the composite signal and the actual mining situation, and reduces the dimensionality of the feature vector, removes characteristic factors that have little influence on the identification result, and verifies the classification result after the model identification is completed to correct the model identification result and improve the classification accuracy of the model.

[0044] Other features and advantages of the embodiments of the present application will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the following detailed description, they are used to explain the embodiments of the present application but do not constitute a limitation on the embodiments of the present application. In the accompanying drawings:

[0046] Figure 1 A flowchart of a method for identifying an excavator working condition according to an embodiment of the present application is schematically shown;

[0047] FIG2( a ) schematically illustrates a bulldozer grading operating condition and a corresponding pressure band diagram according to an embodiment of the present application;

[0048] FIG2( b ) schematically illustrates a soil-leveling working condition and a corresponding pressure band diagram according to an embodiment of the present application;

[0049] Figure 3 The following schematically shows a structural block diagram of an excavator working condition identification device according to an embodiment of the present application. DETAILED DESCRIPTION

[0050] To make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the specific implementation methods described herein are only used to illustrate and explain the embodiments of the present application and are not used to limit the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0051] It should be noted that if the embodiments of the present application involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement status, etc. between the various components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indications will also change accordingly.

[0052] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present application, the descriptions of "first", "second", etc. are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of such features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the fact that they can be implemented by ordinary technicians in this field. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0053] Figure 1 The following schematically shows a flow chart of a method for identifying an excavator working condition according to an embodiment of the present application. Figure 1 As shown, an embodiment of the present application provides a method for identifying an excavator working condition, which is applied to an excavator working condition identification device. The method may include the following steps.

[0054] Step 101: Acquire pilot control signal data and dual pump pressure signal data of an excavator's operating handle;

[0055] Step 102: pre-process the operating handle pilot control signal data and the dual pump pressure signal data to obtain pre-processed data;

[0056] Step 103: construct a feature vector based on the preprocessed data;

[0057] Step 104: Perform feature dimensionality reduction on the feature vector to obtain a reduced-dimensional feature vector;

[0058] Step 105: Input the feature vector after dimension reduction into the classification model to obtain the first working mode of the excavator;

[0059] Step 106: Verify the first operating mode and output a detailed list of operating mode identification results.

[0060] With the development of mechatronics, excavator operating parameters can now be collected and acquired via the controller area network bus. This allows for objectively reflecting current excavator testing requirements, making these parameters an important basis for identifying the excavator's operating phase. In this embodiment, data such as dual pump pressure and the pilot control signal from the operating handle are jointly processed, analyzed, and modeled, resulting in higher accuracy and more convincing identification than with a single signal.

[0061] Specifically, the excavator operating condition identification device can first obtain the excavator's operating handle pilot control signal data and dual pump pressure signal data. The operating handle pilot control signal data includes the dipper arm extension pilot pressure, dipper arm retraction pilot pressure, swing pilot pressure, boom down pilot pressure, boom up pilot pressure, bucket digging pilot pressure, bucket unloading pilot pressure, left travel pilot pressure, and right travel pilot pressure. The dual pump pressure signal data includes the front pump pressure and rear pump pressure. After obtaining the operating handle pilot control signal data and dual pump pressure signal data, the excavator operating condition identification device can preprocess all the data to obtain preprocessed data. Due to the different sensor locations, the various types of data may be time-misaligned. Therefore, the data can be padded (forward-padded and backward-padded with null values) first, and then the padded data can be aligned to obtain aligned data. In addition, because the data signal acquisition is affected by vibration and other interference factors, the signal is often mixed with noise. Therefore, the data signal needs to be filtered. After preprocessing the data, a feature vector can be constructed based on the processed data. Since the number of high-dimensional features can easily interfere with the classification model, thereby affecting the accuracy of the classification model, it is necessary to perform dimensionality reduction on the high-dimensional data. Finally, the feature vector after dimensionality reduction is input into the classification model to identify the first working mode of the excavator. Among them, the first working mode of the excavator refers to the excavation working condition, the leveling working condition and other working modes except the idling working condition, the walking working condition and the crushing working condition. After the working condition is identified by the classification model, the embodiment of the present application can also verify the first working mode to ensure the accuracy of the identification of the excavator working condition, and then output the working condition identification result detailed table.

[0062] Through the above technical solution, the pilot control signal data and dual pump pressure signal data of the operating handle of the excavator are first obtained, and the pilot control signal data and dual pump pressure signal data of the operating handle are preprocessed to obtain preprocessed data. Then, a feature vector is constructed based on the preprocessed data, and the feature vector is subjected to feature dimensionality reduction to obtain a reduced feature vector. Finally, the reduced feature vector is input into the classification model to obtain the first working mode of the excavator, and the first working mode is verified and a detailed table of working condition identification results is output. This application constructs a feature vector based on the basic time domain characteristics of the composite signal and the actual mining situation, and reduces the dimensionality of the feature vector, removes characteristic factors that have little influence on the identification result, and verifies the classification result after the model identification is completed to correct the model identification result and improve the classification accuracy of the model.

[0063] In an embodiment of the present application, the method may further include:

[0064] Obtain the excavator's crushing working mode status, machine construction status, machine travel status and engine speed data;

[0065] The second working mode of the excavator is determined according to the crushing working mode status, machine construction status, machine walking status and engine speed data, and a detailed list of working mode identification results is output.

[0066] In an embodiment of the present application, the second working mode of the excavator includes an idle working mode, a walking working mode and a crushing working mode. In actual applications, when performing excavator working mode identification, the excavator working mode identification device can also obtain the excavator's crushing working mode status, machine construction status, machine walking status and engine speed data. According to the excavator's crushing working mode status, machine construction status, machine walking status and engine speed data, it can be determined whether the excavator is in an idle working mode, a walking working mode and a crushing working mode. Specifically, when the machine construction status is 0 or False, the machine walking status is 0 or False and the engine speed is in a preset range, such as [600, 1100], the excavator working mode identification device determines that the excavator is in an idle working mode. When the machine walking state parameter value is not 0 or False, the excavator working mode identification device determines that the excavator is in a walking working mode. When the crushing working mode status is 1 or True, the excavator working mode identification device determines that the excavator is in a crushing working mode.

[0067] In an embodiment of the present application, the classification model is a libsvm model, and the method may further include:

[0068] Determine the target parameters of the libsvm model to build a classification model.

[0069] In an embodiment of the present application, determining target parameters of the libsvm model to construct a classification model may include:

[0070] Encode the initial parameters of the libsvm model to form an initial population, which includes multiple individuals;

[0071] Determine the fitness of each individual in the initial population separately;

[0072] Determine whether there is an individual whose fitness meets the termination conditions;

[0073] When it is determined that there is an individual whose fitness satisfies the termination condition, the parameters corresponding to the individual whose fitness satisfies the termination condition are determined as the target parameters of the libsvm model;

[0074] When it is determined that there is no individual whose fitness satisfies the termination condition, the initial population is updated by performing selection, crossover and mutation through genetic operators until there is an individual whose fitness satisfies the termination condition.

[0075] In the embodiment of the present application, since the support vector machine (SVM) has good robustness and generalization ability, the libsvm model is used as a classification model to identify the excavator working condition. When constructing a classification model, the penalty parameter c and the kernel parameter g, which have an important influence on the learning ability of the model, are often difficult to determine. The selection of the penalty parameter c can achieve a balance between the model complexity and the training error. The value of the penalty parameter c is generally taken as 1. The parameter g of the kernel function mainly reflects the range characteristics of the training sample data, which directly affects the learning ability of the support vector machine model. It is generally defaulted to 1 / k, where k is the total number of categories. The values ​​of these two parameters have a great influence on the learning ability of the support vector machine model. If the penalty parameter c and the kernel parameter g are too small, the learning ability of the model is poor. If the penalty parameter c and the kernel parameter g are too large, the model will be overfitted. Therefore, in order to improve the learning ability of the model and improve the accuracy of recognition, this application adopts a genetic algorithm to perform parameter optimization within a range. Taking the cross-validation accuracy as the objective function, the optimal parameter values ​​of the penalty parameter c and the kernel parameter g are obtained, so as to determine the target parameters of the libsvm model, that is, the optimal parameters, to build a classification model.

[0076] Specifically, the genetic algorithm encodes parameters c and g, uses the decoded parameter values ​​to train the LibSVM model, and then uses cross-validation to achieve the optimal accuracy. Cross-validation is a common method for selecting and evaluating models in machine learning. The main idea is to use the majority of a given training sample to build a model. A smaller portion of the remaining samples is used to predict the model, and the prediction error of this smaller portion of predictions is calculated. The model with the smallest prediction error is then selected as the optimal model. The genetic algorithm uses the cross-validation accuracy as the fitness value and iterates through genetic operator selection, crossover, and mutation to achieve the optimal parameters for the LibSVM model. In practical applications, the initial parameters of the LibSVM model are first encoded to form an initial population consisting of multiple individuals. The fitness of each individual in the initial population is then determined to determine whether any individual satisfies a termination criterion. If an individual satisfies the termination criterion, the parameters corresponding to the individual are determined as the target parameters for the LibSVM model. If it is determined that there are no individuals whose fitness satisfies the termination condition, the initial population is updated by performing selection, crossover, and mutation through genetic operators until an individual whose fitness satisfies the termination condition exists, and the parameters corresponding to the individual whose fitness satisfies the termination condition are determined as the target parameters of the libsvm model. It should be noted that the technical solution of the present application can use not only genetic algorithms but also simulated annealing algorithms or tabu search algorithms.

[0077] In an embodiment of the present application, preprocessing the operating handle pilot control signal data and the dual pump pressure signal data to obtain preprocessed data may include:

[0078] Performing data filling processing on the pilot control signal data of the operating handle and the dual pump pressure signal data to obtain filled data;

[0079] Performing data alignment processing on the padded data to obtain aligned data;

[0080] The aligned data are subjected to data denoising to obtain preprocessed data.

[0081] In an embodiment of the present application, the operating handle pilot control signal data includes the boom extension pilot pressure, the boom retraction pilot pressure, the rotation pilot pressure, the boom lowering pilot pressure, the boom raising pilot pressure, the bucket excavation pilot pressure, the bucket unloading pilot pressure, the left travel pilot pressure and the right travel pilot pressure. The dual pump pressure signal data includes the front pump pressure and the rear pump pressure. Due to the different positions of the sensors, various types of data may have time misalignment problems. Therefore, the data can be filled first, that is, the data is filled forward and filled backward with empty values, and then the filled data is aligned to obtain aligned data. In addition, since the acquisition of data signals is affected by vibration and other interference factors, the signals are often mixed with noise. Therefore, the data signals also need to be filtered. In an embodiment of the present application, the SG (Savitzky-Golay) algorithm is used to smooth the data. The Savitzky-Golay algorithm is a filtering method based on least squares fitting of local polynomials in the time domain. Its core concept is to perform weighted filtering on the data within a window, where the weights are obtained by least squares fitting a given high-order polynomial. Using the SG algorithm to smooth data can improve the smoothness of the data curve and reduce noise interference. Compared to other denoising algorithms, the SG algorithm can better preserve signal characteristics such as shape and width during filtering and smoothing, thereby more effectively retaining signal change information, thereby making subsequent working condition identification more accurate.

[0082] Figure 2(a) schematically illustrates a bulldozer-type leveling operation and a corresponding pressure band diagram according to an embodiment of the present application; Figure 2(b) schematically illustrates a scoop-type leveling operation and a corresponding pressure band diagram according to an embodiment of the present application. As shown in Figures 2(a) and 2(b), in the embodiment of the present application, the feature vector includes a first feature vector and a second feature vector. Constructing the feature vector based on the preprocessed data may include:

[0083] Get the pressure band of the excavator;

[0084] Intercept the pressure band of the preset size window;

[0085] Extracting the time domain features of the pressure band of a preset size window to obtain a first feature vector;

[0086] The pressure ratios of the excavation signal, the unloading signal, and the rotation signal in the pressure band of the preset size window are extracted to obtain a second eigenvector.

[0087] In this embodiment of the present application, the feature vector includes a first feature vector and a second feature vector. The first feature vector is the time-domain eigenvalue of the excavator's pressure band, and the second feature vector is the pressure contribution of the excavation signal, unloading signal, and rotation signal within the excavator's pressure band. Compared to the prior art, this embodiment of the present application adds the pressure contribution of the excavation signal, unloading signal, and rotation signal within the excavator's pressure band to construct the feature vector, which can help improve the accuracy of model recognition. Specifically, a fixed window method can be used to intercept the pressure band during excavator operation. First, the window size is set to win. In practical applications, it is typically set to 15 seconds per excavation cycle, i.e., win = 150. Then, the time-domain feature parameters of any window are extracted. The time-domain feature parameters include dimensioned and dimensionless time-domain features, including maximum, minimum, mean, standard deviation, root mean square root (RMS), skewness, kurtosis, kurtosis factor, waveform factor, and margin factor. In addition, based on the video and signal data of land leveling, it can be found that there are two main types of excavator land leveling: bulldozing and hooking. As shown in Figure 2(a), bulldozing grading involves using a backhoe to push out soil without any soil in the bucket. This method produces smaller excavation and unloading signals. Another method, as shown in Figure 2(b), involves hooking grading, where soil is present in the bucket. However, because the dipper arm is extended and retracted in one step, the rotary pressure signal is weak, while the dipper arm extension and retraction pressures are significant. Therefore, in addition to the basic time-domain features, this embodiment of the present application also adds the pressure proportions of the excavation signal, unloading signal, and rotary signal in the excavator's pressure band to construct a feature vector, improving the accuracy of model recognition.

[0088] In the embodiment of the present application, performing feature dimensionality reduction on the feature vector to obtain a reduced-dimensional feature vector may include:

[0089] Input the eigenvector and the operating mode label into the random forest function to obtain the contribution rate of each eigenvector;

[0090] Sort the feature vectors according to contribution rate and preset order;

[0091] The preset number of feature vectors is determined as the feature vectors after dimensionality reduction.

[0092] In an embodiment of the present application, there are more than one hundred feature vectors constructed based on the preprocessed data. Since the number of high-dimensional features can easily interfere with the classification model, thereby affecting the accuracy of the classification model, it is necessary to perform dimensionality reduction on the high-dimensional data. Specifically, the random forest algorithm can be used to reduce the dimensionality of the feature vector. The random forest algorithm is one of the classic data mining algorithms. Its core lies in the construction of a decision tree, and through the combination of multiple decision trees, it completes the feature recognition and extraction of sample data. The technical solution of the present application inputs the feature vector and the working mode label into the random forest (RandomForestRegressor) function to obtain the contribution rate of each feature vector, and then inputs a preset number of feature vectors with the highest contribution rate into the classification model for working condition identification, wherein the preset number can be set according to actual conditions, preferably, it can be 20. In this way, reducing the dimensionality of the feature vector by the random forest method can improve the accuracy of classification model recognition.

[0093] In the embodiment of the present application, verifying the first operating mode may include:

[0094] Obtain the peak and trough vibration frequencies of the excavation pilot signal, unloading pilot signal, rotation pilot signal, arm extension pilot signal, and arm retraction pilot signal of the excavator;

[0095] The working condition identification result is verified based on the peak and trough vibration frequencies of the excavation pilot signal, the unloading pilot signal, the rotation pilot signal, the bucket arm extension pilot signal and the bucket arm retraction pilot signal.

[0096] In an embodiment of the present application, after obtaining the first operating condition pattern recognition through the classification model, the recognition results can be further checked and verified. The operating condition recognition results are verified by the peak and trough vibration frequencies of the excavation pilot signal, the unloading pilot signal, the rotation pilot signal, the arm extension pilot signal, and the arm retraction pilot signal. In one example, a maximum value peak_upper and a minimum value peak_low are first set for the excavation pilot signal, the unloading pilot signal, and the rotation pilot signal. Based on the actual situation, two points, p1 and p2, are selected on the pressure bands of the excavation pilot signal and the unloading pilot signal, respectively. When a trough is detected between points p1 and p2 that is less than peak_low, p2 is identified as a valid vibration point. Since the bucket excavation pilot signal and the bucket unloading pilot signal have obvious regular characteristics when the excavator is excavating, if multiple excavation vibration points (i.e., valid vibration points) are identified within a cycle, and the arm extension pilot pressure and arm retraction pilot pressure signals are obvious, it indicates that the excavator is currently in the leveling condition.

[0097] Figure 3The following schematically shows a structural block diagram of an excavator working condition identification device according to an embodiment of the present application. Figure 3 As shown, an embodiment of the present application provides an excavator working condition identification device, which may include:

[0098] Memory 310 configured to store instructions; and

[0099] The processor 320 is configured to call instructions from the memory 310 and implement the above-mentioned method for controlling the boom when executing the instructions.

[0100] Specifically, in the embodiment of the present application, the processor 320 may be configured to:

[0101] Obtain pilot control signal data of the excavator's operating handle and dual pump pressure signal data;

[0102] Preprocessing the operating handle pilot control signal data and the dual pump pressure signal data to obtain preprocessed data;

[0103] Construct feature vectors based on preprocessed data;

[0104] Performing feature dimensionality reduction on the feature vector to obtain a reduced-dimensional feature vector;

[0105] Inputting the reduced-dimensional feature vector into a classification model to obtain a first working mode of the excavator;

[0106] Verify the first operating mode and output a detailed list of operating mode identification results.

[0107] Furthermore, the processor 320 may be configured to:

[0108] Obtain the excavator's crushing working mode status, machine construction status, machine travel status and engine speed data;

[0109] The second working mode of the excavator is determined according to the crushing working mode status, machine construction status, machine walking status and engine speed data, and a detailed list of working mode identification results is output.

[0110] In an embodiment of the present application, the classification model is a libsvm model, and the method further includes:

[0111] Determine the target parameters of the libsvm model to build a classification model.

[0112] Furthermore, the processor 320 may be configured to:

[0113] Encode the initial parameters of the libsvm model to form an initial population, which includes multiple individuals;

[0114] Determine the fitness of each individual in the initial population separately;

[0115] Determine whether there is an individual whose fitness meets the termination conditions;

[0116] When it is determined that there is an individual whose fitness satisfies the termination condition, the parameters corresponding to the individual whose fitness satisfies the termination condition are determined as the target parameters of the libsvm model;

[0117] When it is determined that there is no individual whose fitness satisfies the termination condition, the initial population is updated by performing selection, crossover and mutation through genetic operators until there is an individual whose fitness satisfies the termination condition.

[0118] Furthermore, the processor 320 may be configured to:

[0119] Performing data filling processing on the pilot control signal data of the operating handle and the dual pump pressure signal data to obtain filled data;

[0120] Performing data alignment processing on the padded data to obtain aligned data;

[0121] The aligned data are subjected to data denoising to obtain preprocessed data.

[0122] Furthermore, the processor 320 may be configured to:

[0123] Get the pressure band of the excavator;

[0124] Intercept the pressure band of the preset size window;

[0125] Extracting the time domain features of the pressure band of a preset size window to obtain a first feature vector;

[0126] The pressure ratios of the excavation signal, the unloading signal, and the rotation signal in the pressure band of the preset size window are extracted to obtain a second eigenvector.

[0127] Furthermore, the processor 320 may be configured to:

[0128] Input the eigenvector and the operating mode label into the random forest function to obtain the contribution rate of each eigenvector;

[0129] Sort the feature vectors according to contribution rate and preset order;

[0130] The preset number of feature vectors is determined as the feature vectors after dimensionality reduction.

[0131] Furthermore, the processor 320 may be configured to:

[0132] Obtain the peak and trough vibration frequencies of the excavation pilot signal, unloading pilot signal, rotation pilot signal, arm extension pilot signal, and arm retraction pilot signal of the excavator;

[0133] The working condition identification result is verified based on the peak and trough vibration frequencies of the excavation pilot signal, the unloading pilot signal, the rotation pilot signal, the bucket arm extension pilot signal and the bucket arm retraction pilot signal.

[0134] Through the above technical solution, the pilot control signal data and dual pump pressure signal data of the operating handle of the excavator are first obtained, and the pilot control signal data and dual pump pressure signal data of the operating handle are preprocessed to obtain preprocessed data. Then, a feature vector is constructed based on the preprocessed data, and the feature vector is subjected to feature dimensionality reduction to obtain a reduced feature vector. Finally, the reduced feature vector is input into the classification model to obtain the first working mode of the excavator, and the first working mode is verified and a detailed table of working condition identification results is output. This application constructs a feature vector based on the basic time domain characteristics of the composite signal and the actual mining situation, and reduces the dimensionality of the feature vector, removes characteristic factors that have little influence on the identification result, and verifies the classification result after the model identification is completed to correct the model identification result and improve the classification accuracy of the model.

[0135] An embodiment of the present application also provides a machine-readable storage medium having stored thereon instructions for enabling a machine to execute the above-mentioned method for identifying the working condition of an excavator.

[0136] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0137] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1A device that provides the functions specified in a block or multiple blocks.

[0138] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0139] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0140] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0141] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0142] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0143] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0144] The above are merely embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A method for identifying an excavator working condition, characterized in that: Applied to an excavator working condition identification device, the method includes: Obtain pilot control signal data of the excavator's operating handle and dual pump pressure signal data; Preprocessing the operating handle pilot control signal data and the dual pump pressure signal data to obtain preprocessed data; Constructing a feature vector based on the preprocessed data; Performing feature dimensionality reduction on the feature vector to obtain a reduced-dimensional feature vector; Inputting the dimensionally reduced feature vector into a classification model to obtain a first operating mode of the excavator; Verify the first operating mode and output a detailed list of operating condition identification results; The feature vector includes a first feature vector and a second feature vector, and constructing the feature vector according to the preprocessed data includes: Acquire a pressure band of the excavator; Intercept the pressure band of the preset size window; Extracting the time domain features of the pressure band of the preset size window to obtain a first feature vector; The pressure ratios of the excavation signal, the unloading signal, and the rotation signal in the pressure band of the preset size window are extracted to obtain a second eigenvector.

2. The method according to claim 1, characterized in that The method further comprises: Obtaining data on the crushing working mode, machine operation, machine travel, and engine speed of the excavator; The second working mode of the excavator is determined according to the crushing working mode state, the machine construction state, the machine walking state and the engine speed data, and a detailed list of working condition identification results is output.

3. The method according to claim 1, characterized in that The classification model is a libsvm model, and the method further includes: Determine the target parameters of the libsvm model to build a classification model.

4. The method according to claim 3, characterized in that Determining target parameters of the libsvm model to build a classification model includes: Encoding initial parameters of the libsvm model to form an initial population, wherein the initial population includes a plurality of individuals; Determining the fitness of each individual in the initial population respectively; Determine whether there is an individual whose fitness meets the termination condition; In the case where it is determined that there is an individual whose fitness satisfies the termination condition, determining the parameters corresponding to the individual whose fitness satisfies the termination condition as the target parameters of the libsvm model; When it is determined that there is no individual whose fitness satisfies the termination condition, the initial population is updated by performing selection, crossover and mutation through genetic operators until there is an individual whose fitness satisfies the termination condition.

5. The method according to claim 1, wherein The preprocessing of the operating handle pilot control signal data and the dual pump pressure signal data to obtain preprocessed data includes: Performing data filling processing on the operating handle pilot control signal data and the dual pump pressure signal data to obtain filled data; Performing data alignment processing on the padded data to obtain aligned data; Data denoising is performed on the aligned data to obtain preprocessed data.

6. The method according to claim 1, characterized in that The performing feature dimensionality reduction on the feature vector to obtain a reduced-dimensional feature vector comprises: Inputting the feature vector and the operating mode label into a random forest function to obtain a contribution rate of each feature vector; Sorting the feature vectors according to the contribution rate and a preset order; The preset number of feature vectors is determined as the feature vectors after dimensionality reduction.

7. The method according to claim 1, characterized in that The verifying the first operating mode includes: Acquiring the peak-to-trough vibration frequencies of the excavation pilot signal, the unloading pilot signal, the rotation pilot signal, the bucket arm extension pilot signal, and the bucket arm retraction pilot signal of the excavator; The working condition identification result is verified according to the peak and trough vibration frequencies of the excavation pilot signal, the unloading pilot signal, the rotation pilot signal, the bucket arm extension pilot signal and the bucket arm retraction pilot signal.

8. An excavator working condition identification device, characterized in that: include: a memory configured to store instructions; as well as A processor is configured to call the instructions from the memory and implement the method for identifying an excavator working condition according to any one of claims 1 to 7 when executing the instructions.

9. A machine-readable storage medium, characterized in that The machine-readable storage medium stores instructions for causing a machine to execute the method for identifying an excavator working condition according to any one of claims 1 to 7.

Images