Mining Phase Identification Method, Apparatus, Electronic Device, Storage Medium, and Computer Program Product

Abstract

The present application provides a recognition method, device, electronic device, storage medium, and computer program product for the excavation stage, belonging to the field of computer technology. In this solution, the pilot pressure signal and pump-end pressure signal of the target excavator at the current time step, the operation time of the target excavator at the previous time step, and the excavation stage at the previous time step are obtained; feature extraction is performed on the input vector constructed based on the pilot pressure signal and pump-end pressure signal to obtain a pressure signal feature sequence; according to the pressure signal feature sequence, the operation time at the previous time step, and the excavation stage at the previous time step, the recognition result of the excavation stage of the target excavator at the current time step is determined. By combining the pressure signal feature sequence, the excavation cycle operation stage sequence, and the operation time characteristics, the overall vehicle state is more comprehensively described from three dimensions: the driver's operation intention, the overall vehicle load, and the working condition state, and the final excavation stage recognition result is determined to improve the accuracy of the recognition result.

Description

Technical Field

[0001] The present application relates to the field of computer technologies, and in particular, to a method, device, electronic device, computer-readable storage medium, and computer program product for identifying a digging stage. Background Art

[0002] When the working mode and gear of an existing excavator are determined, the required engine power is a fixed value. However, in fact, the actual power of the excavator changes with different digging cycle stages, and there are often situations where the engine power supply is excessive or insufficient. Adopting power control for the excavator under different working conditions can meet the heavy-load power supply while reducing the waste of energy under light load. This requires being able to accurately identify the working conditions of the excavator in real time.

[0003] However, when identifying the working conditions of an excavator, there are often the following technical problems:

[0004] 1. The existing technology only uses the pump-end pressure signal or the pilot pressure signal alone when identifying the digging cycle stage. The former has hysteresis in the change of pump pressure itself, resulting in the recognition result lagging behind the actual situation; while the pilot pressure only reflects the operation intention of the driver and does not reflect the vehicle load information, which is prone to misidentification (for example, from the digging preparation stage to the digging stage, the vehicle load can reflect whether the excavator encounters materials for digging).

[0005] 2. The existing technology does not merge and compress data. Since the neural network for existing time series recognition (digging cycle recognition mostly uses time series as input vectors) itself has a large amount of computation, the uncompressed data will increase the computation amount, thereby resulting in a high controller load rate.

[0006] 3. The existing technology uses a neural network to identify the digging cycle stage and recognizes the information of the time series involved, but does not consider the dependency relationship between stages and the working time characteristics of each stage, which is prone to misidentification.

[0007] 4. To avoid the occurrence of the gradient explosion phenomenon and also to reduce the computation amount of the neural network model, the input vector of the neural network adopted by the existing technology does not contain information with too long a time length, which also causes the recognition model to be unable to master the time series information of the entire digging cycle, which is also one of the important reasons for misidentification in the digging cycle stage recognition model.

[0008] Therefore, there is an urgent need to develop a new method for identifying the digging stage to solve the current defects and deficiencies. Summary of the Invention

[0009] The content of the present disclosure is partially used to introduce concepts in a brief form, and these concepts will be described in detail in the following detailed implementation section. The content of the present disclosure is not intended to identify the key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0010] Some embodiments of the present disclosure propose a method, apparatus, electronic device, computer-readable storage medium, and computer program product for mining stage identification to at least partially solve the technical problems mentioned in the above background section.

[0011] In a first aspect, some embodiments of the present disclosure provide a method for mining stage identification. The method includes: obtaining a pilot pressure signal and a pump-end pressure signal of a target excavator at a current time step, the operation time of the target excavator at a previous time step, and the mining stage at the previous time step; constructing an input vector based on the pilot pressure signal and the pump-end pressure signal; performing feature extraction on the input vector to obtain a pressure signal feature sequence; and determining an identification result of the mining stage of the target excavator at the current time step based on the pressure signal feature sequence, the operation time at the previous time step, and the mining stage at the previous time step.

[0012] In a second aspect, some embodiments of the present disclosure provide a device for mining stage identification. The device includes: an obtaining unit configured to obtain a pilot pressure signal and a pump-end pressure signal of a target excavator at a current time step, the operation time of the target excavator at a previous time step, and the mining stage at the previous time step; a constructing unit configured to construct an input vector based on the pilot pressure signal and the pump-end pressure signal; an extracting unit configured to perform feature extraction on the input vector to obtain a pressure signal feature sequence; and a determining unit configured to determine an identification result of the mining stage of the target excavator at the current time step based on the pressure signal feature sequence, the operation time at the previous time step, and the mining stage at the previous time step.

[0013] In a third aspect, embodiments of the present application provide an electronic device. The electronic device includes: one or more processors; a storage device for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation manner of the first aspect.

[0014] In a fourth aspect, embodiments of the present application provide a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method described in any implementation manner of the first aspect is implemented.

[0015] Fifth aspect, some embodiments of the present disclosure provide a computer program product, including a computer program which, when executed by a processor, implements the method described in any implementation manner of the above first aspect.

[0016] One embodiment among the various embodiments of the present disclosure has the following beneficial effects: By combining the pressure signal feature sequence, mining the order of the cyclic operation stages, and the operation time characteristics, starting from three dimensions of the driver's operation intention, the vehicle load, and the working condition state, the vehicle state is described more comprehensively, so as to determine the final recognition result of the mining stage and improve the accuracy of the recognition result. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In combination with the accompanying drawings and with reference to the following specific embodiments, the above and other features, advantages, and aspects of the various embodiments of the present disclosure will become more apparent. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the elements and elements are not necessarily drawn to scale.

[0018] Figure 1 is a flowchart of some embodiments of the mining stage recognition method according to the present disclosure;

[0019] Figure 2 is a schematic structural diagram of some embodiments of the mining stage recognition device according to the present disclosure;

[0020] Figure 3 is a schematic structural diagram of an electronic device suitable for implementing some embodiments of the present disclosure.

[0021] Among them, the above-mentioned drawings include the following reference numerals:

[0022] 301, processing device; 302, read-only memory; 303, random access memory; 304, bus; 305, input / output interface; 306, input device; 307, output device; 308, storage device; 309, communication device. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] The embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although some embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. On the contrary, these embodiments are provided to more thoroughly and completely understand the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not used to limit the protection scope of the present disclosure.

[0024] In addition, it should be noted that for the convenience of description, only the parts related to the relevant invention are shown in the drawings. Without conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other.

[0025] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence relationship of the functions performed by these devices, modules or units.

[0026] It should be noted that the modifications of "one" and "multiple" mentioned in the present disclosure are illustrative rather than restrictive. Those skilled in the art should understand that unless otherwise clearly specified in the context, it should be understood as "one or more".

[0027] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only for illustrative purposes, and are not used to limit the scope of these messages or information.

[0028] The present disclosure will be described in detail below with reference to the drawings and in combination with embodiments.

[0029] Reference Figure 1 , shows the flow of some embodiments of the excavation stage identification method according to the present disclosure. The excavation stage identification method includes the following steps:

[0030] Step S101, obtaining the pilot pressure signal and the pump end pressure signal of the target excavator at the current time step, the operation time of the target excavator at the previous time step, and the excavation stage at the previous time step.

[0031] In some embodiments, the execution subject of the excavation stage identification method can obtain the pilot pressure signal and the pump end pressure signal of the target excavator at the current time step, the operation time of the target excavator at the previous time step, and the excavation stage at the previous time step.

[0032] Here, the above-mentioned target excavator generally refers to the excavator whose excavation stage needs to be identified. The current time step generally refers to the time step corresponding to obtaining the above-mentioned pilot pressure signal and pump end pressure signal. The excavation stage at the previous time step generally refers to the excavation stage in which the above-mentioned target excavator is at the previous time step. The operation time at the previous time step generally refers to the operation time of the above-mentioned target excavator in a certain excavation stage at the previous time step.

[0033] Specifically, the pump end pressure generally refers to the pressure at the oil supply outlet of the excavator pump. For example, if the target excavator is supplied with oil by two pumps, the pump end pressure is the general term of the pressure of pump 1 and the pressure of pump 2. In the hydraulic system of the excavator, pump 1 and pump 2 are the oil supply pumps for the actuators of each action of the excavator. Therefore, the pump end pressure can reflect the load condition of the excavator.

[0034] The above-mentioned excavation stage can generally be divided into four stages: excavation preparation stage, excavation stage, return transfer stage, and unloading stage.

[0035] It should be noted that the execution entity electronic device of the excavation stage recognition method can be hardware or software. When the electronic device is hardware, it can be implemented as a distributed cluster composed of multiple servers or terminal devices, or as a single server or a single terminal device. When the electronic device is embodied as software, it can be installed in the above-mentioned hardware devices. It can be implemented as, for example, multiple software or software modules for providing distributed services, or as a single software or software module. No specific limitation is made here.

[0036] Step S102, construct an input vector according to the above-mentioned pilot pressure signal and the above-mentioned pump end pressure signal.

[0037] In some embodiments, the above-mentioned execution entity can construct an input vector according to the above-mentioned pilot pressure signal and the above-mentioned pump end pressure signal.

[0038] As an example, the above-mentioned execution entity can construct an input vector by merging or vectorizing the above-mentioned pilot pressure signal and the above-mentioned pump end pressure signal.

[0039] In some optional implementation manners of some embodiments, the above-mentioned execution entity can perform merge compression processing on the above-mentioned pilot pressure signal and the pump end pressure signal to obtain a merged compression signal;

[0040] Construct an input vector according to the above-mentioned merged compression signal and the following formula:

[0041]

[0042] Where n represents the time series length of the selected input vector, represents the boom pilot pressure set of n sampling points, represents the stick pilot pressure set of n sampling points, represents the bucket pilot pressure set of n sampling points, represents the swing pilot pressure set of n sampling points, represents the pump 1 pressure set of n sampling points, represents the pump 2 pressure set of n sampling points.

[0043] Specifically, the above-mentioned execution entity can perform merge compression processing in the following manner to obtain a merged compression signal:

[0044] The pilot pressure signal and the pump end pressure signal are collected and filtered to remove noise. In addition, in order to reduce the number of input signals and reduce the amount of calculation, the signals are merged and compressed. The bucket retraction and outward turn signals will not occur at the same time, so the bucket retraction and outward turn signals are merged into the bucket signal according to the following formula: , where s is the bucket signal, s1 is the bucket retraction signal, and s2 is the bucket flip signal. Similarly, the boom up and down signals are combined into the boom signal, and the arm retraction and flip signals are combined into the arm signal.

[0045] Afterwards, the above execution body can be constructed as an input vector based on the combined signal, and the expression is as follows:

[0046] ,

[0047] Where n is the time length of the selected input vector, is the boom pilot pressure set of n sampling points, is the set of the arm pilot pressures at n sampling points, is the bucket pilot pressure set of n sampling points, is the set of rotary pilot pressures at n sampling points, is the set of pump 1 pressures at n sampling points, is the pressure set of pump 2 at n sampling points.

[0048] Here, pump end pressure refers to the pressure at the pump oil supply outlet. For example, if the excavator using this solution is supplied by two pumps, the pump end pressure is the general term for the pressure of pump 1 and pump 2. In the hydraulic system of the excavator, pump 1 and pump 2 are the oil supply pumps for the actuators of each action of the excavator. Therefore, the pump end pressure can feedback the load condition of the excavator.

[0049] The pilot pressure signal is combined and compressed to reduce the calculation amount of the model without affecting the model recognition accuracy, thereby reducing the controller load rate, reducing the requirements for the controller, and improving the stability of program operation.

[0050] Step S103, extracting features from the input vector to obtain a pressure signal feature sequence.

[0051] In some embodiments, the execution entity may perform feature extraction on the input vector to obtain a pressure signal feature sequence.

[0052] Here, the pressure signal feature sequence is usually used to characterize the probability that the input vector corresponds to different excavation stages. The pressure signal feature sequence usually includes the probability that the input vector corresponds to the excavation preparation stage, excavation stage, return material transfer stage, and unloading stage.

[0053] As an example, the above-mentioned execution entity can pre-establish a correspondence table between input vectors and pressure signal feature sequences based on a large amount of data, and determine the pressure signal feature sequence corresponding to the above input vector according to the above correspondence table.

[0054] As another example, the above-mentioned execution entity can also perform feature extraction on the above input vector through a pre-trained neural network (such as a bidirectional time series neural network and / or a unidirectional time series network) to obtain a pressure signal feature sequence.

[0055] In some optional implementation manners of some embodiments, the above-mentioned execution entity can use a pre-trained feature extraction model to perform feature extraction on the above input vector to obtain a pressure signal feature sequence, where the above-mentioned feature extraction model is obtained by training a first initial model with a time series neural network as the first initial model, and the above-mentioned first initial model includes an input layer, a time series network processing layer, a hidden layer, a classification layer, and an output layer. After the above-mentioned first initial model is trained, the classification layer and the output layer in the above-mentioned first initial model are discarded to obtain a feature extraction model.

[0056] As an example, the above-mentioned time series neural network generally refers to a neural network structure with time series classification function. The above-mentioned time series neural network can be a unidirectional time series neural network or a bidirectional time series neural network, such as BiRNN, LSTM, BiGRU, Transformer, etc. As an example, the above-mentioned time series network processing layer can be a unidirectional time series network. As another example, the above-mentioned time series network processing layer can also be composed of a bidirectional recurrent layer and a concatenation layer.

[0057] As an example, the above-mentioned execution entity can construct a feature extraction model in the following manner:

[0058] The above-mentioned execution entity can construct a mining stage recognition model based on BiLSTM. The model is mainly divided into the following layers. The combination of FlipLayer, two LSTMLayer, and ConcatenationLayer is used to achieve the purpose of processing data from the forward and reverse directions of the input vector by two LSTMs respectively, so as to achieve the effect of BiLSTM:

[0059] InputLayer: Perform decentralization and normalization processing on the input vector. The formula is as follows:

[0060] ,

[0061] Among them, is the input vector, is the output of InputLayer, is the mean of the training set, is the variance of the training set, is the minimum value of the training set after decentralization, is the maximum value of the training set after decentralization.

[0062] FlipLayer: Flips the vector output by the InputLayer. When the input is , the output is .

[0063] LSTMLayer: Forms a hidden state containing the recognition result based on the data contained in the processed input vector.

[0064] ConcatenationLayer: Concatenates the hidden states output by the forward and backward LSTMLayers together.

[0065] FullConnectedLayer: Reduces the dimension of the features output by the LSTMLayer to 4 features, which can be mapped to the probabilities that the working conditions corresponding to the input belong to 4 excavation cycle stages.

[0066] SoftmaxLayer: A Softmax classifier that forms the probabilities of belonging to different stages at the current moment based on the previously extracted features. The internal specific calculation formula of this network layer will not be elaborated.

[0067] OutputLayer: Based on the output result of the SoftmaxLayer, identifies the stage with the highest probability among the four excavation stages as the current working stage, and calculates the network loss for backpropagation.

[0068] Among them, the results of the SoftmaxLayer and the OutputLayer are only used to evaluate the network loss of the LSTMLayer and support the model training, and are not used for the final identification of the excavation cycle stage.

[0069] During the training process, the time series features extracted by the BiLSTM are limited to 4 dimensions, and a Softmax layer is added for classification. The classification result is used as the model evaluation criterion. Therefore, the extracted features can be mapped to the probabilities that the working conditions corresponding to the input belong to 4 excavation cycle stages.

[0070] For the case where the pressure signal is a long time series and there is a need to further extract effective features. A pre-trained BiLSTM model is used to extract features from the pilot pressure and pump-end pressure signals, and data processing is performed from the forward and reverse directions of the pressure signal respectively to extract the time series features of the pressure signal.

[0071] The advantage of the BiLSTM model in capturing long-term dependencies in time series is fully utilized. The pressure signal is compressed from a two-dimensional feature vector to a one-dimensional feature vector, while the effective time series features in the pressure information are extracted, thereby improving the accuracy of network recognition.

[0072] During the pre-training process of the BiLSTM model, the extracted time series features are set to 4 dimensions and classified using the softmax layer. Finally, the classification is used as the evaluation criterion for the pre-training model, which achieves the effect of giving practical meaning to the extracted features - the extracted features can be mapped to: the probability that the working conditions corresponding to the input belong to the four mining cycle stages. This improves the interpretability of the mining cycle phased recognition model and improves the reliability of the recognition model.

[0073] Step S104, determining the identification result of the excavation stage of the target excavator in the current time step according to the pressure signal feature sequence, the operation time of the previous time step, and the excavation stage of the previous time step.

[0074] In some embodiments, the execution entity may determine the mining stage identification result of the target excavator in the current time step according to the pressure signal feature sequence, the operation time in the previous time step, and the mining stage in the previous time step.

[0075] In some optional implementation methods of some embodiments, the above-mentioned execution entity can input the above-mentioned pressure signal feature sequence, the operation time of the previous time step, and the mining stage of the previous time step into a pre-trained recognition model to obtain the recognition result of the mining stage, wherein the above-mentioned recognition model is obtained by training the above-mentioned second initial model with the DAGSVM model as the second initial model and the radial basis function as the kernel function of the second initial model.

[0076] Specifically, the input of the above DAGSVM is 4 time series feature sequences extracted by the feature extraction model, 2 features of the mining cycle stage to which the previous time step belongs and the time kept in the same mining stage in the previous time step, a total of 6 features.

[0077] These six features are constructed as the input vector of the DAGSVM classifier. The expression of the input vector of the DAGSVM classifier is: ,in, The four time series features extracted by the pre-trained feature extraction model can be mapped as: the probability that the working condition corresponding to the input belongs to the four mining cycle stages, is the mining cycle phase to which the previous time step belongs, is the time spent in the same mining phase during the previous time step.

[0078] Combined with the BiLSTM and SVM structures, the BiLSTM is used to obtain the temporal features of the pressure signal, and the order of the cyclic operation stages and the operation time features are mined and added to jointly construct the input vector of the SVM model, so as to more comprehensively describe the vehicle state from three dimensions: the driver's operation intention, the vehicle load, and the working condition state, and improve the recognition accuracy.

[0079] Among them, the pilot pressure reflects the driver's operation intention, the pump-end pressure can reflect the load information of the excavator, and the order of the cyclic operation stages and the operation time features of the excavation can feedback the working condition of the excavator.

[0080] In some optional implementation manners of some embodiments, the above DAGSVM model includes at least six binary classification nodes, and the above at least six binary classification nodes are set in the directed acyclic graph in the above DAGSVM model according to the following steps:

[0081] Determine the binary classification node with the highest classification accuracy from the above at least six binary classification nodes as the top node in the above directed acyclic graph;

[0082] For the remaining binary classification nodes, determine the relevant nodes with the above top node from the above remaining binary classification nodes, and combine the above relevant nodes in pairs to obtain at least one binary classification node combination. The binary classification node combination with the highest average classification accuracy in the above at least one binary classification node combination is used as the second layer node in the above directed acyclic graph;

[0083] The nodes in the above remaining binary classification nodes except the above second layer nodes are used as the third layer nodes in the above directed acyclic graph.

[0084] Here, the above binary classification node generally refers to a binary classifier. Specifically, the higher the classification accuracy of the binary classification, the easier it is to distinguish the two excavation cycle stages. Therefore, based on the classification accuracy corresponding to each binary classifier, the nodes of the directed acyclic graph are set.

[0085] As an example, the node selection rule is: select the group with the highest classification accuracy as the top node, such as 1VS4. For the two nodes in the second layer, select the groups with the highest average classification accuracy from the two groups of 2VS4 and 1VS3, and 3VS4 and 1VS2 as the left and right two nodes in the second layer, such as 1VS3 and 2VS4. Then, the nodes in the third layer are 1VS2, 2VS3, and 3VS4.

[0086] Due to the advantage that DAGSVM is more suitable for non - linear classification compared with the Softmax layer, the present disclosure adopts DAGSVM as the final classification model, and customizes the DAGSVM classification nodes through the pre - classification results of the binary classifier, improving the applicability of the DAGSVM classifier to the phased recognition of the excavation cycle and further enhancing the accuracy of the excavation cycle recognition model.

[0087] In some optional implementation manners of some embodiments, in response to determining that the above - mentioned excavation stage recognition result satisfies the following rules, the above - mentioned execution entity may correct the above - mentioned excavation stage recognition result according to the corresponding rules and use the corrected result as the excavation stage recognition result:

[0088] When the excavation stage in the previous time step is the excavation preparation stage and the above - mentioned excavation stage recognition result is the excavation stage, if the pilot pressure of the stick outwards turning is greater than the pilot pressure of the bucket inwards turning, correct the above - mentioned excavation stage recognition result to the excavation preparation stage;

[0089] When the excavation stage in the previous time step is the excavation preparation stage and the above - mentioned excavation stage recognition result is not the excavation stage, correct the above - mentioned excavation stage recognition result to the excavation preparation stage;

[0090] When the excavation stage in the previous time step is the excavation stage and the above - mentioned excavation stage recognition result is the transfer stage, if the duration time t_2 in the excavation stage is less than the time threshold t_21, correct the above - mentioned excavation stage recognition result to the excavation stage; if t_2 is greater than or equal to t_21 and less than the time threshold t_22, and the maximum value max1 of the boom raise pilot pressure and the swing pilot pressure is less than the maximum value max2 of the bucket inwards turning and stick inwards turning pilot pressures, correct the above - mentioned excavation stage recognition result to the excavation stage; if t_2 is greater than or equal to t_22, correct the above - mentioned excavation stage recognition result to the transfer stage;

[0091] When the excavation stage in the previous time step is the excavation stage and the above - mentioned excavation stage recognition result is not the transfer stage, correct the above - mentioned excavation stage recognition result to the excavation stage;

[0092] When the excavation stage in the previous time step is the transfer stage and the above - mentioned excavation stage recognition result is the unloading stage, correct the above - mentioned excavation stage recognition result to the unloading stage;

[0093] When the excavation stage in the previous time step is the transfer stage and the above - mentioned excavation stage recognition result is the excavation preparation stage, if the maximum value max3 of the stick outwards turning and bucket outwards turning pilot pressures is greater than the pilot threshold p1, correct the above - mentioned excavation stage recognition result to the unloading stage; if max3 is less than or equal to p1, the corrected result is the transfer stage;

[0094] The mining phase in the previous time step is the transfer phase. When the recognition result of the above mining phase is the mining phase, the recognition result of the above mining phase is corrected to the transfer phase;

[0095] The mining phase in the previous time step is the unloading phase. When the recognition result of the above mining phase is the mining preparation phase, if the continuous time t_4 in the unloading phase is less than the time threshold t_41 and max3 is greater than p1, the recognition result of the above mining phase is corrected to the unloading phase. Otherwise, the recognition result of the above mining phase is corrected to the mining preparation phase;

[0096] The mining phase in the previous time step is the unloading phase. When the recognition result of the above mining phase is not the mining preparation phase, the recognition result of the above mining phase is corrected to the unloading phase.

[0097] By using the correction rule to correct the recognition result, the influence of the model's insufficient grasp of the entire mining cycle information is compensated, and the accuracy and reliability of the recognition result are improved.

[0098] One of the above embodiments of the present disclosure has the following beneficial effects: Combining the pressure signal feature sequence, the order of the mining cycle operation phases, and the operation time characteristics, starting from the three dimensions of the driver's operation intention, the vehicle load, and the working condition state, the vehicle state is described more comprehensively, so as to determine the final recognition result of the mining phase and improve the accuracy of the recognition result.

[0099] Further referring to Figure 2 , as an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a mining phase recognition device. These device embodiments correspond to Figure 1 the method embodiments shown, and the device can be specifically applied to various electronic devices.

[0100] As Figure 2 shown, some embodiments of the mining phase recognition device include: an acquisition unit 201, a construction unit 202, an extraction unit 203, and a determination unit 204. Among them, the acquisition unit 201 is used to acquire the pilot pressure signal and the pump end pressure signal of the target excavator at the current time step, the operation time of the target excavator in the previous time step, and the mining phase in the previous time step; the construction unit 202 is used to construct an input vector according to the pilot pressure signal and the pump end pressure signal; the extraction unit 203 is used to extract features from the input vector to obtain a pressure signal feature sequence; the determination unit 204 is used to determine the recognition result of the mining phase of the target excavator at the current time step according to the pressure signal feature sequence, the operation time in the previous time step, and the mining phase in the previous time step.

[0101] It can be understood that the various units described in the above mining phase recognition device are related to the referenceFigure 1 corresponds to each step in the described method. Thus, the operations, features, and beneficial effects described above for the method also apply to the above-mentioned excavation stage identification device and the units included therein, and will not be elaborated herein.

[0102] One embodiment among the above various embodiments of the present disclosure has the following beneficial effects: By combining the pressure signal feature sequence, the order of the excavation cycle operation stages, and the operation time characteristics, starting from the three dimensions of the driver's operation intention, the vehicle load, and the working condition state, the vehicle state is described more comprehensively, so as to determine the final excavation stage identification result and improve the accuracy of the identification result.

[0103] Reference is made below to Figure 3 , which shows a schematic structural diagram of an electronic device suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is merely an example and should not impose any limitations on the functions and usage scope of the embodiments of the present disclosure.

[0104] As Figure 3 shown, the electronic device may include a processing device 301 (such as a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to the program stored in the read-only memory 302 (ROM) or the program loaded from the storage device 308 into the random access memory 303 (RAM). In the RAM, various programs and data required for the operation of the electronic device are also stored. The processing device 301, the read-only memory 302 (ROM), and the random access memory 303 (RAM) are connected to each other through a bus 304. The input / output interface 305 (I / O interface) is also connected to the bus.

[0105] Generally, the following devices may be connected to the input / output interface 305 (I / O interface): an input device 306 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 308 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 309. The communication device 309 can allow the electronic device to communicate with other devices wirelessly or wiredly to exchange data. Although Figure 3 shows an electronic device having various devices, it should be understood that it is not required to implement or have all the shown devices. More or fewer devices may be alternatively implemented or had. Figure 3 Each block shown in

[0106] In particular, according to some embodiments of the present disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of the present disclosure include a computer program product that includes a computer program carried on a computer-readable medium, and the computer program contains program code for performing the methods shown in the flowcharts. In such some embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by a processing device, the above-mentioned functions defined in the methods of some embodiments of the present disclosure are performed.

[0107] It should be noted that the computer-readable medium in some embodiments of the present disclosure can be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of a computer-readable storage medium can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0108] In some embodiments of the present disclosure, a computer-readable storage medium can be any tangible medium that contains or stores a program, and the program can be used by or in combination with an instruction execution system, apparatus, or device. And in some embodiments of the present disclosure, a computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above.

[0109] A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, and the computer-readable signal medium can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted by any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0110] In some embodiments, the client and the server can communicate using any currently known or future-developed network protocol such as HTTP (HyperText Transfer Protocol), and can be interconnected with digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include local area networks ("LAN"), wide area networks ("WAN"), the Internet (e.g., the Internet), and end-to-end networks (e.g., ad hoc end-to-end networks), as well as any currently known or future-developed networks.

[0111] The above computer-readable medium can be included in the above electronic device; or can exist separately without being assembled into the electronic device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by the electronic device, the electronic device is caused to: obtain the pilot pressure signal and the pump-end pressure signal of the target excavator at the current time step, the operation time of the target excavator at the previous time step and the excavation stage at the previous time step; construct an input vector according to the pilot pressure signal and the pump-end pressure signal; perform feature extraction on the input vector to obtain a pressure signal feature sequence; determine the excavation stage recognition result of the target excavator at the current time step according to the pressure signal feature sequence, the operation time at the previous time step, and the excavation stage at the previous time step.

[0112] Computer program code for performing the operations of some embodiments of the present disclosure can be written in one or more programming languages or combinations thereof. The above programming languages include object-oriented programming languages - such as Java, Smalltalk, C++; and also include conventional procedural programming languages - such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network - including a local area network (LAN) or a wide area network (WAN) - or can be connected to an external computer (e.g., by using an Internet service provider to connect through the Internet).

[0113] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a part of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, as well as combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0114] The units described in some embodiments of the present disclosure can be implemented in software or in hardware. The described units can also be provided in a processor. For example, it can be described as: a processor includes an acquisition unit, a construction unit, an extraction unit, and a determination unit. Among them, the names of these units do not constitute a limitation to the unit itself in some cases. For example, the acquisition unit can also be described as "the unit that acquires the pilot pressure signal and the pump-end pressure signal of the target excavator at the current time step, the operation time of the target excavator at the previous time step, and the excavation stage at the previous time step".

[0115] The functions described above herein can be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that can be used include: Field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), Systems on Chip (SOCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0116] Some embodiments of the present disclosure also provide a computer program product, including a computer program that, when executed by a processor, implements any of the above-described excavation stage identification methods.

[0117] The above description is only some preferred embodiments of the present disclosure and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by the specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the technical solutions formed by mutually replacing the above features with the technical features (but not limited to) disclosed in the embodiments of the present disclosure that have similar functions.

Claims

1. A method for identifying a mining stage, characterized in that, Including: Obtain the pilot pressure signal and pump-end pressure signal of the target excavator at the current time step, the operation time of the target excavator at the previous time step, and the excavation stage at the previous time step; The excavation stage is divided into 4 excavation cycle stages: excavation preparation stage, excavation stage, swing transfer stage, and unloading stage; Construct an input vector according to the pilot pressure signal and the pump-end pressure signal; Perform feature extraction on the input vector to obtain a pressure signal feature sequence; the pressure signal feature sequence characterizes the probabilities of the input vector corresponding to different excavation stages; Input the pressure signal feature sequence, the operation time at the previous time step, and the excavation stage at the previous time step into a pre-trained recognition model to determine the excavation stage recognition result of the target excavator at the current time step. Among them, the recognition model uses the DAGSVM model as the second initial model and trains the second initial model with the radial basis function as the kernel function of the second initial model. And the input of the DAGSVM model is: 4 time series feature sequences of the pressure signal feature sequence, the excavation cycle stage to which the previous time step belongs, and 2 features of the time staying in the same excavation stage at the previous time step, a total of 6 features. Among them, the 4 time series feature sequences are mapped to the probabilities of the working conditions corresponding to the input vector belonging to the 4 excavation cycle stages.

2. The mining stage identification method according to claim 1, wherein Constructing an input vector according to the pilot pressure signal and the pump-end pressure signal includes: Perform filtering and denoising on the pilot pressure signal and the pump-end pressure signal; Perform merging and compression processing on the filtered and denoised pilot pressure signals and pump-end pressure signals of the same type to obtain a merged and compressed signal; Construct an input vector according to the merged and compressed signal and the following formula: ， where n represents the time series length of the selected input vector, represents the boom pilot pressure set of n sampling points, represents the stick pilot pressure set of n sampling points, represents the bucket pilot pressure set of n sampling points, represents the swing pilot pressure set of n sampling points, represents the pump 1 pressure set of n sampling points, represents the pump 2 pressure set of n sampling points.

3. The mining stage identification method according to claim 1, characterized in that Performing feature extraction on the input vector to obtain a pressure signal feature sequence includes: Use a pre-trained feature extraction model to perform feature extraction on the input vector to obtain a pressure signal feature sequence. Among them, the feature extraction model uses a time series neural network as the first initial model and trains the first initial model. Among them, the first initial model includes an input layer, a time series network processing layer, a hidden layer, a classification layer, and an output layer. After the first initial model is trained, discard the classification layer and the output layer in the first initial model to obtain a feature extraction model.

4. The mining stage identification method according to claim 1, wherein The DAGSVM model includes at least six binary classification nodes. Set the at least six binary classification nodes in the directed acyclic graph in the DAGSVM model according to the following steps, including: Determine the binary classification node with the highest classification accuracy among the at least six binary classification nodes as the top node in the directed acyclic graph; For the remaining binary classification nodes, determine the relevant nodes with respect to the top node from the remaining binary classification nodes, and combine the relevant nodes in pairs to obtain at least one binary classification node combination. Take the binary classification node combination with the highest average classification accuracy in the at least one binary classification node combination as the second layer node in the directed acyclic graph; Nodes among the remaining binary classification nodes other than the second-layer nodes are used as third-layer nodes in the directed acyclic graph.

5. The mining phase identification method according to claim 1, characterized in that The mining phase identification method further includes: In response to determining that the mining phase identification result satisfies the following rules, the mining phase identification result is corrected according to the corresponding rules, and the corrected result is used as the mining phase identification result: When the mining phase in the previous time step is the mining preparation phase and the mining phase identification result is the mining phase, if the pilot pressure of the stick out is greater than the pilot pressure of the bucket in, the mining phase identification result is corrected to the mining preparation phase; When the mining phase in the previous time step is the mining preparation phase and the mining phase identification result is not the mining phase, the mining phase identification result is corrected to the mining preparation phase; When the mining phase in the previous time step is the mining phase and the mining phase identification result is the transfer and unloading phase, if the continuous time t_2 in the mining phase is less than the time threshold t_21, the mining phase identification result is corrected to the mining phase; when t_2 is greater than or equal to t_21, less than the time threshold t_22, and the maximum value max1 of the boom up pilot pressure and the swing pilot pressure is less than the maximum value max2 of the bucket in and the stick in pilot pressure, the mining phase identification result is corrected to the mining phase; if t_2 is greater than or equal to t_22, the mining phase identification result is corrected to the transfer and unloading phase; When the mining phase in the previous time step is the mining phase and the mining phase identification result is not the transfer and unloading phase, the mining phase identification result is corrected to the mining phase; When the mining phase in the previous time step is the transfer and unloading phase and the mining phase identification result is the unloading phase, the mining phase identification result is corrected to the unloading phase; When the mining phase in the previous time step is the transfer and unloading phase and the mining phase identification result is the mining preparation phase, if the maximum value max3 of the pilot pressure of the stick out and the bucket out is greater than the pilot threshold p1, the mining phase identification result is corrected to the unloading phase; if max3 is less than or equal to p1, the corrected result is the transfer and unloading phase; When the mining phase in the previous time step is the transfer and unloading phase and the mining phase identification result is the mining phase, the mining phase identification result is corrected to the transfer and unloading phase; When the mining phase in the previous time step is the unloading phase and the mining phase identification result is the mining preparation phase, if the continuous time t_4 in the unloading phase is less than the time threshold t_41 and max3 is greater than p1, the mining phase identification result is corrected to the unloading phase; otherwise, the mining phase identification result is corrected to the mining preparation phase; When the mining phase in the previous time step is the unloading phase and the mining phase identification result is not the mining preparation phase, the mining phase identification result is corrected to the unloading phase.

6. A device for identifying the excavation stage, characterized in that, including: An acquisition unit, configured to acquire a pilot pressure signal and a pump-end pressure signal of a target excavator at a current time step, the operation time of the target excavator at a previous time step, and the excavation stage at the previous time step; the excavation stage is divided into four excavation cycle stages: an excavation preparation stage, an excavation stage, a slewing and transfer stage, and a discharging stage; A construction unit, configured to construct an input vector according to the pilot pressure signal and the pump-end pressure signal; An extraction unit, configured to perform feature extraction on the input vector to obtain a pressure signal feature sequence; the pressure signal feature sequence represents the probabilities of the input vector corresponding to different excavation stages; A determination unit, configured to input the pressure signal feature sequence, the operation time at the previous time step, and the excavation stage at the previous time step into a pre-trained recognition model to determine the excavation stage recognition result of the target excavator at the current time step, wherein the recognition model uses a DAGSVM model as a second initial model and uses a radial basis function as a kernel function of the second initial model to train the second initial model, and the input of the DAGSVM model is: 4 time series feature sequences of the pressure signal feature sequence, the excavation cycle stage to which the previous time step belongs, and 2 features of the time staying in the same excavation stage in the previous time step, a total of 6 features, wherein the 4 time series feature sequences are mapped to the probabilities of the working conditions corresponding to the input vector belonging to the 4 excavation cycle stages; 7. An electronic device, characterized in that, Comprising: One or more processors; A storage device, on which one or more programs are stored, When the one or more programs are executed by the one or more processors, the one or more processors are enabled to implement the excavation stage recognition method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, On which a computer program is stored, wherein the computer program, when executed by a processor, is capable of implementing the excavation stage recognition method according to any one of claims 1 to 5.

9. A computer program product, characterized in that, Comprising a computer program, which, when executed by a processor, is capable of implementing the excavation stage recognition method according to any one of claims 1 to 5.

Images