Method, device, equipment for determining working object of working machinery, and working machinery

Abstract

The present invention provides a method, device, equipment and operating machinery for determining the operating object of an operating machinery, which belongs to the technical field of intelligent operating machinery. The method collects the working parameters of the operating machinery through a CAN bus; inputs the working parameters into a working object determination model, and outputs the operating object of the operating machinery, wherein the working object determination model is obtained after training based on predetermined working parameter samples and working object samples, directly obtains the working parameters of the operating machinery itself through the CAN bus, and then determines the operating object according to its own working parameters, without the need for additional sensors, effectively saving the cost of determining the operating object.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent working machinery, and in particular to a method, device, equipment and working machinery for determining a working object of a working machinery. Background Art

[0002] Currently, operating machinery primarily uses visual recognition or cylinder pressure identification to identify the work object. Visual recognition primarily uses convolutional neural networks to visually identify the surface of the work object, while cylinder pressure identification primarily relies on selecting a threshold.

[0003] However, both visual recognition and cylinder pressure recognition require additional dedicated sensors, and the recognition cost is relatively high. Summary of the Invention

[0004] The present invention provides a method, device, equipment and working machine for determining a working object of a working machine, so as to solve the defect in the prior art that working object identification requires the addition of additional sensors at high cost, and realizes the determination of the working object through the working machine's own parameters, thereby effectively reducing the cost of determining the working object.

[0005] The present invention provides a method for determining an operating object of an operating machine, comprising:

[0006] Collect the working parameters of the operating machinery through the CAN bus;

[0007] The working parameters are input into a working object determination model, and the working object of the working machine is output, wherein the working object determination model is obtained after training based on predetermined working parameter samples and working object samples.

[0008] According to a method for determining a work object of a work machine provided by the present invention, the training process of the work object determination model includes:

[0009] Collecting sample data of working parameters of the working machinery and sample data of the working objects;

[0010] Preprocessing the working parameter sample data to identify valid sample data in the working parameter sample data, where the valid data is data corresponding to when the working machine is working;

[0011] The valid sample data and the job object sample data are used to train an initial machine learning model to construct a job object determination model.

[0012] According to a method for determining a working object of a working machine provided by the present invention, after identifying valid sample data in the working parameter sample data, the method further includes:

[0013] Extract time characteristic series data of valid sample data within a preset time length;

[0014] The statistics of each valid data sample in each of the time feature series data are determined, and the statistics are used as new valid sample data.

[0015] According to a method for determining a working object of a working machine provided by the present invention, the statistical quantity includes at least one of a mean value, a variance, a covariance, a time length, a slope, and a kurtosis.

[0016] According to the present invention, a method for determining a working object of a working machine further includes:

[0017] Collect new working parameter data and new working object data of new machine types;

[0018] The work object determination model is generalized to the new type of work machine according to the new working parameter data and the new work object data.

[0019] According to a method for determining a working object of a working machine provided by the present invention, the working parameters include a first main pump pressure, a first main pump flow, a second main pump pressure, a second main pump flow, engine parameters and hydraulic system parameters.

[0020] The present invention further provides a device for determining an operating object of an operating machine, comprising:

[0021] The acquisition module is used to collect the working parameters of the operating machinery through the CAN bus;

[0022] The determination module is used to input the working parameters into a working object determination model and output the working object of the working machine, wherein the working object determination model is obtained after training based on predetermined working parameter samples and working object samples.

[0023] The present invention also provides an electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for determining an operating object of an operating machine as described above is implemented.

[0024] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for determining an operating object of an operating machine as described in any one of the above is implemented.

[0025] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for determining a working object of a working machine as described above is implemented.

[0026] The present invention also provides a working machine, including the working machine used to execute the working object determination method of the working machine as described in any one of the above items.

[0027] The present invention provides a method, device, equipment and working machine for determining the working object of a working machine. The method collects the working parameters of the working machine through a CAN bus; inputs the working parameters into a working object determination model, and outputs the working object of the working machine, wherein the working object determination model is obtained after training based on predetermined working parameter samples and working object samples, and directly obtains the working parameters of the working machine itself through the CAN bus, and then determines the working object according to its own working parameters, without the need for additional sensors, effectively saving the cost of determining the working object. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0029] Figure 1 It is a flow chart of a method for determining an operating object of an operating machine provided by the present invention;

[0030] Figure 2 It is a structural schematic diagram of the working object determination device of the working machine provided by the present invention;

[0031] Figure 3 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0032] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0033] The following combination Figure 1-Figure 3 The present invention describes a method, device, equipment and working machine for determining a working object of a working machine.

[0034] Figure 1 It is a flow chart of the method for determining the working object of the working machine provided by the present invention.

[0035] like Figure 1As shown, an embodiment of the present invention provides a method for determining an operating object of an operating machine, the execution subject of which may be a vehicle-mounted control system, and mainly includes the following steps:

[0036] 101. Collect the working parameters of the operating machinery through the CAN bus.

[0037] In a specific implementation, we'll use an excavator as an example to illustrate the excavator's operating tasks, which include earthwork, rockwork, and so on. CAN, short for Controller Area Network (CAN), is one of the most widely used fieldbuses and a standard bus for control systems and embedded industrial control area networks. CAN communication meets diverse requirements for safety, comfort, convenience, low power consumption, and low cost.

[0038] Various operating parameters of the working machine, i.e., the excavator, are collected via the CAN bus. These parameters include the pressure and flow of the excavator's first and second main pumps, various engine operating parameters, and various hydraulic system operating parameters. The engine operating parameters include engine speed and torque, while the hydraulic system operating parameters include hydraulic oil temperature. The specific implementation process includes various operating parameters of the working machine itself that can be obtained via the CAN bus, and no further examples are provided.

[0039] 102. Input working parameters into a working object determination model, and output a working object of the working machine, wherein the working object determination model is obtained through training based on predetermined working parameter samples and working object samples.

[0040] Specifically, after collecting the machine's operating parameters via the CAN bus, they are fed into the work object determination model. The model then uses internal calculations to determine the work object corresponding to the currently acquired parameters. For excavators, these work objects include earthwork, stonework, and other different work objects. The work object determination model is trained using pre-determined working parameter and work object samples, and uses a machine learning model to determine work objects, making it more intelligent.

[0041] This embodiment provides a method for determining the working object of a working machine, which collects the working parameters of the working machine through a CAN bus; inputs the working parameters into a working object determination model, and outputs the working object of the working machine, wherein the working object determination model is obtained after training based on predetermined working parameter samples and working object samples, directly obtains the working parameters of the working machine itself through the CAN bus, and then determines the working object based on its own working parameters, without the need for additional sensors, effectively saving the cost of determining the working object.

[0042] Furthermore, this embodiment explains the training process of the work object determination model, which mainly includes: collecting working parameter sample data and work object sample data of the working machinery; preprocessing the working parameter sample data to identify valid sample data in the working parameter sample data, where the valid data is the data corresponding to the working machinery when performing the operation; using the valid sample data and the work object sample data to train the initial machine learning model to construct a work object determination model.

[0043] Specifically, data collection, or the acquisition of training samples, is required first. Similar to model application, data from the operating machinery, specifically the excavator, is collected via the CAN bus. This involves reading the excavator data through its CAN port. Labels can be added to the data, allowing for supervised learning in subsequent machine learning model training. These labels can also be added during post-processing or through a built-in program in the excavator controller while the excavator is operating. The collected data includes various operating parameters and corresponding excavator operation objects, meaning each operating parameter sample corresponds to a type of operation object.

[0044] After data collection is complete, the operating parameter sample data needs to be preprocessed. This primarily involves removing noise and abnormal data, filtering out data from abnormal operating conditions, such as deleting data related to the excavator's automatic idling. At the same time, single excavation data is filtered using the bucket and arm pilot pressures. For each excavation action, in addition to manually applying pilot pressure for pattern recognition, a time series-based algorithm model can also be trained to determine the starting point of the excavation action. The valid sample data obtained is the data corresponding to the excavator's excavation operation. Excavation operation refers to the process from the excavator bucket contacting the work object to the excavation of the work object, which does not include the rotation of the excavator's various mechanical arms.

[0045] After preprocessing and identifying the data, the valid sample data with the greatest correlation with the mining operation is obtained. The valid sample data and the operation object sample data are then input into the initial machine learning model for training. The parameters of the initial machine learning model are continuously adjusted to obtain the final operation object determination model.

[0046] After building the work object determination model, the work object determination model can also be verified, that is, a certain amount of earthwork and stone excavation data is collected again to verify the accuracy of the classification model. The F1 score is used as an indicator, which must reach at least 0.9. At this time, the amount of test data collected is at least 10% of the training data, thereby ensuring the accuracy of the work object determination model.

[0047] Furthermore, based on the above embodiment, after identifying valid sample data in the working parameter sample data, this embodiment may further include: extracting time feature sequence data of the valid sample data within a preset duration, where the preset duration is set based on the duration of a single operation, and of course, it may also be the average duration of multiple operations, etc., and can be determined according to the user's own needs during the specific implementation process. In this embodiment, the single operation duration of the operating machine is preferably used as the preset duration; determining the statistics of each valid data sample in each time feature sequence data, and using the statistics as new valid sample data. The statistics include: at least one of: mean, variance, covariance, time duration, slope, and kurtosis.

[0048] Specifically, extracting the time characteristic sequence data of the effective sample data within the preset time length refers to taking the data within a certain time period as a sample. For example, taking the engine speed as an example, extracting the time characteristic sequence data of the effective sample data within the preset time length refers to extracting all the engine speeds of the excavator during the excavation action, not the engine speed at a certain moment. For example, if the excavation action duration of the excavator is 6s, then all changes in the engine speed of the excavator within 6s are obtained. The same applies to other parameters, thereby obtaining the time characteristic sequence data of each working parameter within the preset time length.

[0049] After extracting the time feature series data, statistics are performed on each valid data sample within each time feature series, and the statistics are used as the latest valid sample data. The specific statistics can be at least one of the statistical mean, variance, covariance, time length, slope, and kurtosis. For example, each time feature series data is regarded as a sample, and each sample is the average value of the parameter or other statistics within a preset time length. It is not the value at a specific moment, but the statistical value of the excavator during an excavation action.

[0050] For each data mining run, each statistic is considered a new feature. All parameters are then normalized, and data quality and integrity are ensured before entering the initial machine learning model. Finally, a statistical learning algorithm (which requires extensive feature engineering) or a deep learning algorithm (which requires numerous hyperparameter presetting) is employed to repeatedly adjust the model parameters and build a model specific to the job.

[0051] Furthermore, based on the above embodiments, the work object determination model can be deployed in the cloud. The data algorithm developed based on traditional statistical learning is easy to interpret and deploy, and can be deployed on the cloud big data platform to monitor the excavator's work objects in real time, thereby completing a large number of calculation processes of the work object determination model in the cloud, reducing the transportation pressure of the work machinery itself. The work machinery itself only needs to perform simple data remote communication transmission, thereby better realizing remote monitoring of the work objects of the work machinery and better realizing intelligent control of the work machinery.

[0052] Furthermore, based on the above embodiment, this embodiment may also include: collecting new working parameter data and new working object data of the new model working machinery; and generalizing the working object determination model to the new model working machinery based on the new working parameter data and the new working object data. Since each working machinery has its own working parameters that may vary to a certain extent, the working object determination models trained for different working machinery may also have slight differences. In order to reduce the time for model training, it is possible to choose to directly generalize the already trained working object determination model to the new model, thereby reducing the length of time required to train the neural network model for the new model. This allows each working machinery to have a corresponding working object determination model, thereby improving the accuracy of working object determination.

[0053] Specifically, when the model needs to be generalized, earthwork and stone excavation data from other machine types is collected. For data standard labels, the architecture and parameters of the trained neural network model are transferred and trained on the data of the new machine type, thereby obtaining a work object determination model applicable to other new machine types. At this time, the parameter weights of the latter several layers are retrained and adjusted to generalize the model to more machine types. Generalization refers to the ability of a neural network to obtain reasonable output for data not encountered during training. In other words, the concept of learning from certain data and correctly applying the acquired knowledge to other data is called generalization. Therefore, it is no longer necessary to train the neural network model separately for each type of work machine; only generalization is required.

[0054] Based on the same general inventive concept, the present application also protects a device for determining an operating object of an operating machine. The device for determining an operating object of an operating machine provided by the present invention is described below. The device for determining an operating object of an operating machine described below and the method for determining an operating object of an operating machine described above can be referred to each other.

[0055] Figure 2 It is a structural schematic diagram of the working object determination device of the working machine provided by the present invention.

[0056] like Figure 2 As shown, an embodiment of the present invention provides a device for determining a working object of a working machine, comprising:

[0057] The acquisition module 201 is used to acquire the working parameters of the operating machine via the CAN bus;

[0058] The determination module 202 is used to input the working parameters into a working object determination model and output the working object of the working machine, wherein the working object determination model is obtained after training based on predetermined working parameter samples and working object samples.

[0059] This embodiment provides a device for determining an operating object of an operating machine, which collects the operating parameters of the operating machine through a CAN bus; inputs the operating parameters into an operating object determination model, and outputs the operating object of the operating machine, wherein the operating object determination model is obtained after training based on predetermined working parameter samples and working object samples, directly obtains the operating parameters of the operating machine itself through the CAN bus, and then determines the operating object based on its own working parameters, without the need for additional sensors, effectively saving the cost of determining the operating object.

[0060] Furthermore, this embodiment also includes a model building module for:

[0061] Collecting sample data of working parameters of the working machinery and sample data of the working objects;

[0062] Preprocessing the working parameter sample data to identify valid sample data in the working parameter sample data, where the valid data is data corresponding to when the working machine is working;

[0063] The valid sample data and the job object sample data are used to train an initial machine learning model to construct a job object determination model.

[0064] Furthermore, this embodiment also includes a model building module, which is specifically used to:

[0065] Extract time characteristic series data of valid sample data within a preset time length;

[0066] The statistics of each valid data sample in each of the time feature series data are determined, and the statistics are used as new valid sample data.

[0067] Furthermore, the statistical quantity in this embodiment includes at least one of mean, variance, covariance, time length, slope and kurtosis.

[0068] Furthermore, this embodiment includes a migration module for:

[0069] Collect new working parameter data and new working object data of new machine types;

[0070] The work object determination model is generalized to the new type of work machine according to the new working parameter data and the new work object data.

[0071] Furthermore, the operating parameters in this embodiment include the first main pump pressure, the first main pump flow, the second main pump pressure, the second main pump flow, engine parameters and hydraulic system parameters.

[0072] Based on the same general inventive concept, the present invention also protects a working machine, which is used to execute the working object determination method of the working machine of any of the above embodiments, and the working machine includes machines such as excavators.

[0073] Figure 3 It is a structural schematic diagram of the electronic device provided by the present invention.

[0074] like Figure 3 As shown, the electronic device may include: a processor 310, a communications interface 320, a memory 330, and a communications bus 340. The processor 310, the communications interface 320, and the memory 330 communicate with each other via the communications bus 340. The processor 310 may invoke logic instructions in the memory 330 to execute a method for determining an operating object of an operating machine. The method includes: collecting operating parameters of the operating machine via a CAN bus; inputting the operating parameters into an operating object determination model; and outputting the operating object of the operating machine. The operating object determination model is obtained through training based on predetermined operating parameter samples and operating object samples.

[0075] In addition, the logic instructions in the above-mentioned memory 330 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0076] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the working object determination method of the working machinery provided by the above methods, the method including: collecting the working parameters of the working machinery through the CAN bus; inputting the working parameters into the working object determination model, and outputting the working object of the working machinery, wherein the working object determination model is obtained after training based on predetermined working parameter samples and working object samples.

[0077] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, it is implemented to execute the working object determination method of the working machinery provided by the above-mentioned methods. The method includes: collecting the working parameters of the working machinery through the CAN bus; inputting the working parameters into the working object determination model, and outputting the working object of the working machinery, wherein the working object determination model is obtained after training based on predetermined working parameter samples and working object samples.

[0078] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0079] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiments.

[0080] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for determining a working object of a working machine, characterized in that: include: Collect the working parameters of the operating machine through the CAN bus; Inputting the working parameters into a working object determination model and outputting the working object of the working machine, wherein the working object determination model is obtained after training based on predetermined working parameter samples and working object samples; The training process of the job object determination model includes: Collecting sample data of the working parameters of the working machine itself and sample data of the working object; Preprocessing the working parameter sample data to identify valid sample data in the working parameter sample data, wherein the valid sample data is data corresponding to the working machine when the working machine is working, wherein the preprocessing includes screening out abnormal working condition data; Using the valid sample data and the job object sample data to train an initial machine learning model to construct a job object determination model; After identifying the valid sample data in the working parameter sample data, the method further includes: Extract time characteristic series data of valid sample data within a preset time length; The statistics of the valid sample data in each of the time feature series data are determined, and the statistics are used as new valid sample data.

2. The method for determining a work object of a work machine according to claim 1, wherein: The statistical quantity includes at least one of mean value, variance, covariance, time length, slope and kurtosis.

3. The method for determining a work object of a work machine according to any one of claims 1 to 2, characterized in that: Also includes: Collect new working parameter data and new working object data of new machine types; The work object determination model is generalized to the new type of work machine according to the new working parameter data and the new work object data.

4. The method for determining a work object of a work machine according to claim 1, wherein: The operating parameters include a first main pump pressure, a first main pump flow, a second main pump pressure, a second main pump flow, engine parameters, and hydraulic system parameters.

5. A device for determining an operating object of an operating machine, characterized in that: include: The acquisition module is used to collect the working parameters of the operating machine through the CAN bus; a determination module, configured to input the working parameters into a working object determination model and output a working object of the working machine, wherein the working object determination model is obtained by training based on predetermined working parameter samples and working object samples; The training process of the job object determination model includes: Collecting sample data of the working parameters of the working machine itself and sample data of the working object; Preprocessing the working parameter sample data to identify valid sample data in the working parameter sample data, wherein the valid sample data is data corresponding to the working machine when the working machine is working, wherein the preprocessing includes screening out abnormal working condition data; Using the valid sample data and the job object sample data to train an initial machine learning model to construct a job object determination model; After identifying the valid sample data in the working parameter sample data, the method further includes: Extract time characteristic series data of valid sample data within a preset time length; The statistics of the valid sample data in each of the time feature series data are determined, and the statistics are used as new valid sample data.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the work object determination method of the work machine according to any one of claims 1 to 4 is implemented.

7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for determining a work object of a work machine according to any one of claims 1 to 4 is implemented.

8. A working machine, characterized in that: The working machine is configured to execute the working object determination method of the working machine according to any one of claims 1 to 4.

Images