A load forecasting method, apparatus, equipment, program product, and engineering machinery.

Abstract

This application provides a load prediction method, apparatus, equipment, program product, and construction machinery. The method includes: acquiring current state detection information of the working mechanism, current pressure detection information of the working mechanism, and current terrain detection information of the excavation area; determining the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information and the current state detection information; and determining the pressure prediction information of the working mechanism at the next moment based on the terrain prediction information, the state prediction information, and the current pressure detection information, wherein the pressure prediction information is used to determine the load of the construction machinery. This application not only enables real-time load prediction of construction machinery but also eliminates the need to establish complex soil or material models, reducing computational complexity, improving the real-time performance of predictions, and conforming to the complex and ever-changing working conditions of excavating machinery, thus possessing wide applicability.

Description

Technical Field

[0001] This application relates to the field of engineering machinery technology, specifically to a load prediction method, device, equipment, program product, and engineering machinery. Background Technology

[0002] When excavators and other construction machinery are performing excavation operations, their load conditions directly affect the stability and energy efficiency of the entire machine. Real-time and accurate prediction of the load on construction machinery is of vital importance for assessing the instantaneous power output of the entire machine, planning energy-efficient excavation trajectories, and designing fully automated excavation control strategies.

[0003] Limited by sensor technology and complex and ever-changing operating environments, the load on excavating machinery exhibits strong nonlinearity and uncertainty, posing significant challenges to the stable control and efficient operation of the entire machine. Existing methods for predicting excavating machinery load have many limitations and shortcomings in engineering applications, and the industry urgently needs a new load prediction method that can predict the load of excavating machinery in real time and accurately. Summary of the Invention

[0004] In view of this, the embodiments of this application aim to provide a load prediction method, apparatus, equipment, program product and engineering machinery that can predict the load force of excavating machinery in real time and accurately.

[0005] According to a first aspect of the embodiments of this application, a load prediction method is provided, applied to construction machinery, the construction machinery including a working mechanism for performing excavation operations, the method comprising:

[0006] The current status detection information of the working mechanism, the current pressure detection information of the working mechanism, and the current terrain detection information of the excavation area are obtained. The current status detection information includes the pose detection value and / or velocity detection value of the working mechanism at the current moment.

[0007] Based on the current terrain detection information and the current status detection information, determine the terrain prediction information of the excavation area and the status prediction information of the working mechanism at the next moment;

[0008] Based on the terrain prediction information, the state prediction information, and the current pressure detection information, the pressure prediction information of the working mechanism at the next moment is determined;

[0009] Based on the pressure prediction information of the operating mechanism at the next moment, determine the load prediction value of the operating mechanism at the next moment.

[0010] Optionally, determining the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information and the current state detection information includes:

[0011] Based on the current terrain detection information, the preset target terrain information, the current status detection information, and the historical status detection information of the working mechanism within the past preset time period, the terrain prediction information of the excavation area and the status prediction information of the working mechanism at the next moment are determined.

[0012] Optionally, determining the pressure prediction information of the operating mechanism at the next moment based on the terrain prediction information, the state prediction information, and the current pressure detection information includes:

[0013] Based on the amount of terrain change between the current terrain detection information and the terrain prediction information, and the amount of state change between the current state detection information and the state prediction information, a first pressure change is determined, wherein the first pressure change represents the predicted pressure change of the working mechanism between the current moment and the next moment.

[0014] Based on the first pressure change and the current pressure detection information, the pressure prediction information of the working mechanism at the next moment is determined.

[0015] Optionally, determining a first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, and the state change between the current state detection information and the state prediction information, and determining the pressure prediction information of the working mechanism at the next moment based on the first pressure change and the current pressure detection information, includes:

[0016] The current terrain detection information, the terrain prediction information, the current state detection information, the state prediction information, and the current pressure detection information are input into a pre-trained first pressure prediction model. The first pressure prediction model determines a first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, and the state change between the current state detection information and the state prediction information. Based on the first pressure change and the current pressure detection information, the first pressure prediction model determines the pressure prediction information of the working mechanism at the next moment.

[0017] Optionally, the first pressure prediction model is a time series neural network model, and the step of determining the first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, and the state change between the current state detection information and the state prediction information, includes:

[0018] Based on the terrain change between the current terrain detection information and the terrain prediction information, the state change between the current state detection information and the state prediction information, and the hidden state at the previous moment, the first pressure change and the hidden state at the current moment are determined. The hidden state at the previous moment includes the model input information within a preset time period in the past and the intermediate state information corresponding to the model input information.

[0019] Optionally, determining the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information and the current state detection information, and determining the pressure prediction information of the working mechanism at the next moment based on the terrain prediction information, the state prediction information, and the current pressure detection information, includes:

[0020] The current state detection information, the current pressure detection information, and the current terrain detection information are input into a pre-trained second pressure prediction model to obtain the pressure prediction information of the working mechanism at the next moment. The second pressure prediction model is a time series neural network model, which includes a first sub-module and a second sub-module.

[0021] The first submodule is used to determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism in the next moment based on the current terrain detection information, the current state detection information, the preset target terrain information, and the historical state detection information within the past preset time period in the hidden state of the previous moment.

[0022] The second submodule is used to determine the first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, the state change between the current state detection information and the state prediction information, and the hidden state at the previous moment, and to determine the pressure prediction information of the working mechanism at the next moment based on the first pressure change and the current pressure detection information. The hidden state at the previous moment includes the model input information within a preset time period in the past and the intermediate state information corresponding to the model input information.

[0023] According to a second aspect of the embodiments of this application, a load prediction device is provided, applied to construction machinery, the construction machinery including a working mechanism for performing excavation operations, the device comprising:

[0024] The data acquisition unit is used to acquire the current status detection information of the working mechanism, the current pressure detection information of the working mechanism, and the current terrain detection information of the excavation area. The current status detection information includes the pose detection value and / or velocity detection value of the working mechanism at the current moment.

[0025] The first prediction unit is used to determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information and the current state detection information.

[0026] The second prediction unit is used to determine the pressure prediction information of the working mechanism at the next moment based on the terrain prediction information, the state prediction information and the current pressure detection information;

[0027] The load prediction unit is used to determine the load prediction value of the working mechanism at the next moment based on the pressure prediction information of the working mechanism at the next moment.

[0028] According to a third aspect of the embodiments of this application, a computer program product is provided, including computer program instructions that, when executed by a processor, cause the processor to implement the load prediction method as described in any one of the first aspects of the embodiments of this application.

[0029] According to a fourth aspect of the embodiments of this application, a load prediction device is provided, including a memory and a processor;

[0030] The memory is connected to the processor and is used to store programs;

[0031] The processor is configured to implement the load prediction method as described in any one of the first aspects of the embodiments of this application by running a program in the memory.

[0032] According to a fifth aspect of the embodiments of this application, an engineering machine is provided, comprising:

[0033] The operating mechanism, the detection equipment, and the load prediction device as described in the third aspect of the embodiments of this application, wherein,

[0034] The detection equipment includes a first detection device, a second detection device, and a third detection device, and the first detection device, the second detection device, and the third detection device are respectively communicatively connected to the load prediction device.

[0035] The working mechanism is used to perform excavation operations;

[0036] The first detection device is used to collect the current status detection value of the working mechanism in real time;

[0037] The second detection device is used to collect the current pressure detection value of the operating mechanism in real time;

[0038] The third detection device is used to collect terrain information of the area to be excavated in real time.

[0039] Optionally, the construction machinery is an excavator;

[0040] The working mechanism includes a boom, a stick, and a bucket;

[0041] The first detection equipment includes boom condition detection equipment, stick condition detection equipment, and bucket condition detection equipment;

[0042] The second testing equipment includes boom pressure testing equipment, stick pressure testing equipment, and bucket pressure testing equipment.

[0043] The load prediction method provided in this application first acquires the current status detection information of the working mechanism, the current pressure detection information of the working mechanism, and the current terrain detection information of the excavation area in real time; then, based on the current terrain detection information and the current status detection information, it determines the terrain prediction information of the excavation area and the status prediction information of the working mechanism at the next moment; finally, based on the terrain prediction information, the status prediction information, and the current pressure detection information, it determines the pressure prediction information of the working mechanism at the next moment.

[0044] The load prediction method provided in this application has two advantages. First, by acquiring the current status detection information of the working mechanism, the current pressure detection information of the working mechanism, and the current terrain detection information of the excavation area in real time, it can continuously and accurately predict the pressure prediction information of the working mechanism at the next moment, and thus predict the load situation of the working mechanism at the next moment, meeting the real-time and accuracy requirements of excavating machinery for load prediction. Second, it eliminates the need to establish complex soil and material models, reducing the complexity of calculation and improving the real-time performance of prediction. Moreover, it does not depend on specific soil and material parameters, reducing prediction errors caused by changes in soil or material types. It conforms to the complex and ever-changing actual working conditions of excavating machinery and has wider applicability. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0046] Figure 1 This is a structural schematic diagram of an engineering machine provided in an embodiment of this application.

[0047] Figure 2 This is a flowchart illustrating a load prediction method provided in an embodiment of this application.

[0048] Figure 3 This is a schematic diagram of a processing flow for determining the pressure prediction information of the operating mechanism at the next moment, provided in an embodiment of this application.

[0049] Figure 4 This is a schematic diagram of another processing flow for determining the pressure prediction information of the operating mechanism at the next moment, provided as an embodiment of this application.

[0050] Figure 5 This is a schematic diagram of a long short-term neural network provided in an embodiment of this application.

[0051] Figure 6 This is a flowchart illustrating another load prediction method provided in an embodiment of this application.

[0052] Figure 7 This is a schematic diagram of the structure of a load prediction device provided in an embodiment of this application.

[0053] Figure 8 This is a schematic diagram of the structure of a load prediction device provided in an embodiment of this application. Detailed Implementation

[0054] The technical solution of this application is applicable to various scenarios requiring load prediction for excavating machinery, such as mining, infrastructure construction, earthwork engineering, and river dredging. In these scenarios, it is typically necessary to understand the load status of the excavating machinery in real time and adjust operating strategies accordingly to prevent overloading or inefficient operation, thereby ensuring operational efficiency, machinery safety, and cost control. The technical solution of this application can continuously and accurately predict the pressure forecast information of the operating mechanism at the next moment, thus predicting the load status of the construction machinery at the next moment, meeting the real-time and accuracy requirements of excavating machinery load prediction.

[0055] The technical solutions provided in this application can be applied, by way of example, to hardware devices such as processors, electronic devices, and servers (including cloud servers), or packaged into software programs for execution. When the hardware device executes the processing procedure of the technical solutions in this application, or when the aforementioned software program is run, the target task can be automatically split and the application programming interfaces required by the task can be automatically invoked to achieve the purpose of the target task. This application only provides illustrative descriptions of the specific processing procedure of the technical solutions in this application and does not limit the specific implementation form of the technical solutions in this application. Any technical implementation form that can execute the processing procedure of the technical solutions in this application can be adopted by this application.

[0056] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0057] Before introducing the solution proposed in this application, the relevant technologies will first be introduced:

[0058] When excavators and other construction machinery are performing excavation operations, their load conditions directly affect the stability and energy efficiency of the entire machine. Real-time and accurate prediction of the load on construction machinery is of vital importance for assessing the instantaneous power output of the entire machine, planning energy-efficient excavation trajectories, and designing fully automated excavation control strategies.

[0059] Limited by sensor technology and complex, ever-changing operating environments, the load on excavating machinery exhibits strong nonlinearity and uncertainty, posing significant challenges to the stable control and efficient operation of the entire machine. Existing methods for predicting excavating machinery load have many limitations and shortcomings in engineering applications. For example, while the discrete element method and finite element method offer high simulation accuracy, their computational overhead involves numerous numerical iterations, severely limiting their application in real-time controllers. Analytical expression methods, although computationally efficient, are overly dependent on material property parameters such as density, hardness, moisture content, and composition. In the real world, material properties vary greatly and are non-uniform, requiring parameter calibration and identification for each type of material. All of these factors hinder the engineering application of traditional excavating machinery load prediction methods, and the industry urgently needs a new load prediction method capable of predicting excavating machinery load in real time and accurately.

[0060] In view of this, the embodiments of this application are committed to providing a load prediction method, device, equipment, program product and engineering machinery that can predict the load force of excavating machinery in real time and accurately, and will be described in detail in the following embodiments.

[0061] Exemplary construction machinery

[0062] To facilitate understanding, the implementation environment of the load prediction method provided in this application embodiment will first be described exemplarily. Please refer to [link / reference]. Figure 1 , Figure 1 This is a structural schematic diagram of an engineering machinery, and the load prediction method provided in this application can be applied to this engineering machinery as an example.

[0063] like Figure 1 As shown, the engineering machinery includes a first testing device 110, a second testing device 120, a third testing device 130, a working mechanism 140, and a load prediction device 150.

[0064] The first detection device 110, the second detection device 120, and the third detection device 130 are each communicatively connected to the load prediction device 150. It should be noted that the communication connection can be a wireless communication connection, a wired communication connection, or any combination thereof; it can be a direct communication connection, an indirect communication connection, or any combination thereof; this application does not limit the specific type of connection.

[0065] The working mechanism 140 is the part of the construction machinery that directly participates in the excavation operation and is used to perform the excavation work. Optionally, the working mechanism 140 includes a first component and a second component, wherein the second component is the part that directly contacts the material and performs the excavation, and the first component is connected between the main body of the construction machinery and the second component to support and operate the second component.

[0066] The first detection device 110 is used to collect the current state detection values ​​of the working mechanism in real time, such as the pose detection values ​​related to the pose of the working mechanism at the current moment, the speed detection values ​​related to the moving speed of the working mechanism at the current moment, etc., and send the collected current state detection values ​​to the load prediction device 150 through the communication network, so that the load prediction device 150 can obtain the current state detection values ​​of the working mechanism. Optionally, the first detection device 110 includes at least one pose sensor (e.g., inertial measurement unit (IMU), lidar, depth camera, camera, ultrasonic sensor, angle sensor, etc.) and / or at least one speed sensor (e.g., photoelectric encoder, Doppler radar, or laser velocimeter, etc.).

[0067] The second detection device 120 is used to collect the current pressure detection value of the working mechanism in real time, and send the collected current pressure detection value to the load prediction device 150 through a communication network. The load prediction device 150 can then obtain the current status detection value of the working mechanism. Optionally, the second detection device 120 can be various types of pressure detection devices such as pressure sensors, pressure transmitters, and pressure detection modules built into hydraulic systems.

[0068] The third detection device 130 is used to collect terrain information of the current excavation area in real time, and send the collected terrain information to the load prediction device 150 through a communication network. The load prediction device 150 can then obtain the current status detection value of the working mechanism. Optionally, the third detection device 130 can be various types of terrain sensing and measurement devices such as lidar, depth camera, and ultrasonic sensor array.

[0069] It should be noted that the type, quantity, and location of the first testing device 110, the second testing device 120, and the third testing device 130 are determined based on factors such as the required testing accuracy, the complexity of the operating mechanism, and the budget. This application does not impose any restrictions on these factors.

[0070] The load prediction device 150 is responsible for executing the load prediction method provided in this application. Specifically, it first acquires the current state detection information of the working mechanism, the current pressure detection information of the working mechanism, and the current terrain detection information of the excavation area. The current state detection information includes the pose detection value and / or speed detection value of the working mechanism at the current moment. Then, based on the current terrain detection information and the current state detection information, it determines the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment. Finally, based on the terrain prediction information, the state prediction information, and the current pressure detection information, it determines the pressure prediction information of the working mechanism at the next moment. The pressure prediction information is used to determine the load of the construction machinery. Other specific implementations of the load prediction method will be elaborated in subsequent content and will not be detailed here.

[0071] The load prediction device 150 can be various types of processing devices, such as a standalone processor or integrated into other controllers, such as a vehicle controller. The specific type of the load prediction device 150 is determined according to the application scenario and performance requirements, and this application does not limit it.

[0072] As an optional implementation, the construction machinery is an excavator. The working mechanism includes a boom, a stick, and a bucket. The first detection equipment includes boom status detection equipment, stick status detection equipment, and bucket status detection equipment. The second detection equipment includes boom pressure detection equipment, stick pressure detection equipment, and bucket pressure detection equipment.

[0073] Exemplary methods

[0074] Figure 2 This is a flowchart illustrating a load prediction method provided in an embodiment of this application. Figure 2 As shown, the load prediction method provided in this embodiment can be applied to [specific applications]. Figure 1 The construction machinery in the process is specifically executed by the load prediction equipment within the construction machinery, including steps S201-S204:

[0075] S201. Obtain the current status detection information of the working mechanism, the current pressure detection information of the working mechanism, and the current terrain detection information of the excavation area. The current status detection information includes the pose detection value and / or velocity detection value of the working mechanism at the current moment.

[0076] The current status detection information of the working mechanism can be understood as the actual operating status information of the working mechanism of the construction machinery at the current moment.

[0077] Optionally, a first detection device is installed on or near the construction machinery to collect the status parameters of the machinery's operating mechanism in real time and send the collected status parameters to a load prediction device. The load prediction device determines the current status detection information of the operating mechanism based on the received status parameters of the operating mechanism at the current moment.

[0078] The current state detection information includes the pose detection value and / or velocity detection value of the working mechanism at the current moment.

[0079] Optionally, the construction machinery is equipped with a pose detection device that can directly detect the pose value of the working mechanism at the current moment.

[0080] Optionally, the construction machinery is equipped with a pose detection device that can detect pose parameters related to the pose of the working mechanism. Based on the pose parameters detected at the current moment and the pre-set and stored structural parameters of the construction machinery, the pose detection value of the working mechanism at the current moment is calculated.

[0081] For example, the construction machinery is an excavator, and the posture detection device includes a boom angle sensor, a stick angle sensor, and a bucket angle sensor. The posture detection device collects the angles of the boom, stick, and bucket in real time, and, combined with the pre-set and stored excavator structural parameters, can calculate the posture detection values ​​of the boom, stick, and bucket at the current moment.

[0082] Optionally, the construction machinery is equipped with speed sensors (such as encoders, tachogenerators, etc.) that can directly detect the speed of the working mechanism at the current moment, or detect speed parameters related to the current speed of the working mechanism, in order to calculate the speed of the working mechanism at the current moment. For example, the construction machinery is an excavator, and the posture detection device includes a boom speed sensor, a stick speed sensor, and a bucket speed sensor, capable of collecting the speeds of the boom, stick, and bucket in real time.

[0083] It should be noted that the current status detection information may also include other information related to the operating status of the working mechanism, such as acceleration information and temperature information, which will be determined according to the actual application scenario and requirements.

[0084] The current pressure detection information of the operating mechanism can be understood as the actual pressure value that the operating mechanism bears or generates at the current moment.

[0085] Optionally, the mechanical equipment adopts a hydraulic system, and the current pressure detection information of the working mechanism can be understood as the hydraulic force of the working mechanism at the current moment.

[0086] Optionally, the construction machinery is equipped with a second detection device (such as a pressure sensor) that can detect the pressure information of the working mechanism in real time.

[0087] For example, the construction machinery is an excavator using a hydraulic system. The second detection device includes a boom pressure sensor, a stick pressure sensor, and a bucket pressure sensor. The current pressure detection information of the working structure includes the hydraulic pressure of the boom hydraulic cylinder, the hydraulic pressure of the stick hydraulic cylinder, and the hydraulic pressure of the bucket hydraulic cylinder.

[0088] The status and pressure monitoring information of the operating mechanism are dynamically changing, which can reflect the working status of the operating mechanism in real time and provide an important basis for subsequent load prediction.

[0089] The current terrain detection information of the excavation area can be understood as the terrain feature data of the excavation area corresponding to the construction machinery at the current moment, which covers the terrain conditions in the current excavation area, such as slope, depressions, and ground protrusions.

[0090] Optionally, a third detection device (such as lidar, binocular camera, etc.) installed on the construction machinery can be used to scan or photograph the excavation area in real time. Invalid information (such as trees, rocks, step boundaries, excavator's own working devices, etc.) in the data collected by the sensors can be filtered out by point cloud or visual semantic segmentation technology to generate a three-dimensional terrain model of the current excavation area of ​​the construction machinery, thereby obtaining the terrain feature data of the current excavation area, that is, the current terrain detection information of the excavation area.

[0091] Alternatively, other methods and existing technologies can be used to obtain the current terrain detection information of the excavation area. For example, drones equipped with high-definition cameras or video cameras can be used to take aerial photos to obtain the current terrain detection information of the excavation area. Alternatively, the Global Positioning System (GPS) can be used to obtain the precise location information of the excavation area, and combined with the database of the Geographic Information System (GIS) to obtain the current terrain detection information of the excavation area.

[0092] Furthermore, in actual work, the excavation area of ​​construction machinery may be relatively wide, while the working mechanism of construction machinery usually only operates within a specific working plane at a certain moment. Therefore, when analyzing the current terrain detection information of the excavation area, more attention should be paid to the terrain information of the plane where the working mechanism is located at the current moment.

[0093] As an optional implementation method, the current terrain detection information of the mining area can be understood as the terrain feature data of the working plane corresponding to the construction machinery at the current moment.

[0094] Optionally, after acquiring the 3D terrain data of the current excavation area of ​​the construction machinery, terrain feature data corresponding to the current working plane can be filtered from the 3D terrain model of the current excavation area. This process can reduce the amount of data processed and analyzed, effectively improve the overall execution efficiency of load prediction, and save computing power.

[0095] Optionally, the current working plane of the construction machinery can be determined based on the current working position and posture of the working mechanism.

[0096] Optionally, the current working plane of the construction machinery can be determined by measuring the rotation angle of the construction machinery's rotating platform and / or the pre-planned digging angle.

[0097] S202. Based on the current terrain detection information and the current state detection information, determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment.

[0098] In actual operation, the terrain information of the excavation area is a dynamically changing parameter that constantly changes with the continuous excavation work of the construction machinery. Specifically, the excavation action of the working mechanism at each moment will cause an immediate change in the terrain of the excavation area. This change is based on the terrain of the excavation area at the previous moment and the real-time excavation action of the working mechanism, which is directly related to the current state information of the working mechanism (such as key parameters such as position, attitude, speed, and acceleration). Therefore, the terrain information of the excavation area at the current moment and the current state information of the working mechanism are directly related to the terrain information of the excavation area at the next moment. By comprehensively considering the terrain information of the excavation area at the current moment and the current state information of the working mechanism, it is possible to predict the terrain information of the excavation area at the next moment.

[0099] In actual operation, the status information of the working mechanism is also a dynamically changing parameter, influenced by the terrain information of the excavation area and the status information of the working mechanism at the previous moment. Specifically, the terrain information of the excavation area provides the environmental conditions for the working mechanism's movement, determining the resistance and support it may encounter during movement. For example, if the current terrain is steep, the working mechanism may encounter greater resistance in the next moment, thus affecting its operating status. The status information of the working mechanism at the previous moment reflects its movement capability under the current terrain. The status change of the working mechanism is continuous, and the current status information provides the basis for the status information at the next moment. Therefore, the current terrain information of the excavation area and the current status information of the working mechanism directly relate to the status information of the working mechanism at the next moment. By comprehensively considering the current terrain information of the excavation area and the current status information of the working mechanism, it is possible to predict the status information of the working mechanism at the next moment.

[0100] As an optional implementation, step S202 is implemented based on the processing concept of a neural network model. In this implementation, step S202 includes: inputting the current terrain detection information of the excavation area and the current state detection information of the working mechanism into a pre-trained first prediction model, and the first prediction model determines the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information of the excavation area and the current state detection information of the working mechanism.

[0101] Specifically, in this embodiment, a first prediction model is pre-trained. Optionally, the first prediction model can be obtained by predictive training on a first training sample, which includes the sample's current terrain detection information, sample's current state detection information, and label information (the labeled terrain prediction information of the excavation area at the next moment and the labeled state prediction information of the working mechanism at the next moment). The first training aims to enable the first prediction model to determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the sample's current terrain detection information and sample's current state detection information in the first training sample, and to achieve a loss function determined by the terrain prediction information and state prediction information output by the model and the labeled terrain prediction information and labeled state prediction information in the label information that meets preset requirements. The trained first prediction model can capture the complex relationship between the terrain information of the excavation area and the state of the working mechanism, and then determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information of the excavation area and the current state detection information of the working mechanism.

[0102] Optionally, the first prediction model can be implemented based on various deep learning neural network structures, such as CNN (Convolutional Neural Networks), RNN (Recurrent Neural Networks), CRNN (Convolutional Recurrent Neural Network), LSTM (Long Short-Term Memory), GRU (Gated Recurrent Unit), etc., and this application does not limit it.

[0103] Optionally, step S202 can also be implemented based on one or more of the following processing concepts: physical model, time series prediction, rule prediction, simulation prediction, etc. This application does not limit this.

[0104] S203. Based on the terrain prediction information, the state prediction information, and the current pressure detection information, determine the pressure prediction information of the working mechanism at the next moment.

[0105] When a working mechanism is performing excavation work, its stress state is a dynamic and continuously changing process. The stress on the mechanism at any given moment is not isolated but closely related to its previous stress state. That is, the pressure on the mechanism at a particular moment evolves from the pressure at the previous moment and is influenced by the pressure information from the previous moment. Specifically, due to the inertia of an object and the continuity of forces, the pressure on the mechanism at the current moment will continue to exert its effect at the next moment unless other external factors significantly alter this stress state. For example, if the mechanism is currently under significant pressure, then due to inertia and the continuity of forces, it is likely to still experience significant pressure at the next moment, unless the terrain suddenly changes or the mechanism suddenly stops moving.

[0106] The terrain information of the excavation area provides the external conditions of the operating environment for construction machinery. There is an interaction between the construction machinery and the terrain of the excavation area. Terrain features (such as undulation and slope) affect the stress state of the construction machinery. The terrain prediction information of the excavation area at the next moment provides the terrain features that the construction machinery will face at the next moment, which can help predict the pressure information of the construction machinery's operating mechanism at the next moment. For example, if the terrain prediction information of the excavation area at the next moment indicates that the terrain will become more rugged, then the operating mechanism may need to withstand greater resistance, resulting in increased pressure.

[0107] The status information of the working mechanism affects how it interacts with the excavated material and the surrounding environment, revealing its interaction with terrain features and thus influencing the stress information experienced by the working mechanism. The next-moment status prediction information provides a forecast of the working mechanism's operational state, helping to predict the stress information of the construction machinery's working mechanism at the next moment. For example, if the next-moment status prediction indicates that the working mechanism will perform faster or more vigorous excavation movements, it may need to withstand greater stress.

[0108] As can be seen from the above, the terrain prediction information of the excavation area at the next moment, the status prediction information of the working mechanism at the next moment, and the current pressure detection information of the working mechanism will all affect the pressure information of the working mechanism at the next moment. Combining these three types of information and considering their joint effects and mutual influences can better predict the pressure information of the working mechanism at the next moment.

[0109] Specifically, after obtaining the terrain prediction information of the excavation area at the next moment, the status prediction information of the working mechanism at the next moment, and the current pressure detection information of the working mechanism, the first strategy is executed to comprehensively analyze the terrain prediction information of the excavation area, the status prediction information of the working mechanism, and the current pressure detection information of the working mechanism, so as to determine the pressure prediction information of the working mechanism at the next moment.

[0110] As an optional implementation, the first strategy can be implemented based on the processing concept of dynamic change analysis, including: determining a first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, and the state change between the current state detection information and the state prediction information, whereby the first pressure change represents the predicted pressure change of the operating mechanism between the current moment and the next moment; and determining the pressure prediction information of the operating mechanism at the next moment based on the first pressure change and the current pressure detection information. This will be elaborated upon in detail later and will not be repeated here.

[0111] As an optional implementation, the first strategy can also be implemented based on the processing concept of a neural network model. Optionally, step S203 includes: inputting the terrain prediction information, the state prediction information, and the current pressure detection information into a pre-trained second prediction model, and having the second prediction model determine the pressure prediction information of the working mechanism at the next moment based on the terrain prediction information, the state prediction information, and the current pressure detection information.

[0112] Specifically, in this embodiment, a second prediction model is pre-trained. Optionally, the second prediction model is obtained by performing a first pressure prediction training on a second training sample. The second training sample includes terrain information at a second time step, state information at a second time step, and pressure information at a first time step, wherein the second time step is the time step following the first time step, and the sample label of the second training sample is the pressure information labeled at the second time step. The first pressure prediction training aims to enable the second prediction model to predict the pressure information at the second time step based on the terrain information, state information, and pressure information at the second time step, and to ensure that the loss function value determined by the model-predicted pressure information and the labeled pressure information at the second time step meets preset requirements. The trained second prediction model can predict the pressure information at the second time step based on the input terrain information, state information, and pressure information at the second time step.

[0113] The second prediction model can be implemented based on various deep learning neural network structures, such as CNN (Convolutional Neural Networks), RNN (Recurrent Neural Networks), CRNN (Convolutional Recurrent Neural Networks), etc., and this application does not limit it.

[0114] S204. Based on the pressure prediction information of the working mechanism at the next moment, determine the load prediction value of the working mechanism at the next moment.

[0115] The pressure prediction information determined in step S203 is crucial for determining the load on the operating mechanism. Accurately predicting the load on the operating mechanism can optimize operational efficiency, reduce energy consumption, extend equipment lifespan, and ensure operational safety. As for the specific method for determining the load on the operating mechanism based on its pressure information, relevant existing technologies can be referenced, and this application does not limit this approach.

[0116] As an optional implementation, the load prediction value of the working mechanism at the next moment is determined based on the pressure information of the working mechanism, including: determining the driving force prediction value of the working mechanism at the next moment based on the pressure prediction information of the working mechanism at the next moment, and using the driving force prediction value to represent the load status of the working mechanism at the next moment.

[0117] The driving force of the working mechanism can be understood as the force generated by the actuator (such as a hydraulic cylinder) of the working mechanism during the working process to move the load, while the pressure of the working mechanism is the key factor in generating the driving force.

[0118] Specifically, after determining the pressure prediction information of the working mechanism at the next moment, the predicted driving force of the working mechanism at the next moment can be determined by calculating based on the pressure prediction information and the geometric parameters of the actuator of the working mechanism. Taking the working mechanism using a hydraulic cylinder as an example, the driving force of the working mechanism is the hydraulic force, which is equal to the pressure multiplied by the effective working area of ​​the piston of the hydraulic cylinder.

[0119] Furthermore, when the working mechanism performs a specific action, the working state of the working mechanism actuator will change, thereby affecting the effective working area of ​​the working mechanism actuator (such as the effective working area of ​​the piston of a hydraulic cylinder). Therefore, before determining the predicted value of the driving force of the working mechanism at the next moment based on the pressure prediction information and the geometric parameters of the working mechanism actuator, it is necessary to first determine the working state of the working mechanism actuator (such as whether the rod chamber of the hydraulic cylinder is compressed or the rodless chamber is compressed) based on the attitude information in the state prediction information of the working mechanism at the next moment, and then determine the geometric parameters of the working mechanism actuator (such as the effective working area of ​​the first piston when the rod chamber is compressed or the effective working area of ​​the second piston when the rodless chamber is compressed).

[0120] The load prediction method provided in this application first acquires the current status detection information of the working mechanism, the current pressure detection information of the working mechanism, and the current terrain detection information of the excavation area in real time; then, based on the current terrain detection information and the current status detection information, it determines the terrain prediction information of the excavation area and the status prediction information of the working mechanism at the next moment; finally, based on the terrain prediction information, the status prediction information, and the current pressure detection information, it determines the pressure prediction information of the working mechanism at the next moment.

[0121] The load prediction method provided in this application has two advantages. First, by acquiring the current status detection information of the working mechanism, the current pressure detection information of the working mechanism, and the current terrain detection information of the excavation area in real time, it can continuously and accurately predict the pressure prediction information of the working mechanism at the next moment, and thus predict the load situation of the working mechanism at the next moment, meeting the real-time and accuracy requirements of excavating machinery for load prediction. Second, it eliminates the need to establish complex soil and material models, reducing the complexity of calculation and improving the real-time performance of prediction. Moreover, it does not depend on specific soil and material parameters, reducing prediction errors caused by changes in soil or material types. It conforms to the complex and ever-changing actual working conditions of excavating machinery and has wider applicability.

[0122] As an optional implementation, after determining the pressure prediction information for the next moment, the weld stress of the engineering machinery at the next moment can also be determined based on the pressure prediction information for the next moment. As for the specific implementation method of determining the weld stress of the engineering machinery based on the pressure information of the working mechanism, it can be implemented with reference to relevant existing technologies, and this application does not limit it in this regard.

[0123] As an optional implementation, step S202, "determining the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information and the current state detection information," includes: determining the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information, preset target terrain information, the current state detection information, and the historical state detection information of the working mechanism within a preset time period.

[0124] Specifically, after obtaining the current status detection information of the working mechanism and the current terrain detection information of the excavation area, the system further combines the preset target terrain information and the stored historical status detection information of the working mechanism to comprehensively determine the terrain prediction information of the excavation area and the status prediction information of the working mechanism at the next moment.

[0125] The current terrain detection information is real-time acquired terrain data, reflecting the immediate condition of the excavation area. The current status detection information is real-time acquired operational status information, reflecting the immediate condition of the operating mechanism. The preset target terrain information is the pre-set final target terrain state for this excavation operation, i.e., the terrain state expected to be achieved after the excavation is completed. The target terrain information provides the direction and objective of this excavation operation. By comparing it with the current terrain detection information, the possible excavation direction required to reach the target terrain can be predicted. The historical status information reflects the status changes of the operating mechanism over a preset period of time, reflecting the movement trend and operating mode of the operating mechanism. For example, if the operating mechanism has been excavating at a certain speed for a period of time, it can help predict whether the operating mechanism will continue to excavate at that speed in the near future. Combining the above four aspects of information allows for a more accurate prediction of the terrain of the excavation area and the status of the operating mechanism at the next moment.

[0126] This approach can also be implemented based on the processing concept of neural network models.

[0127] Optionally, a third prediction model is pre-trained. Current terrain detection information, preset target terrain information, current state detection information, and historical state detection information are input into the pre-trained third prediction model. The third prediction model outputs terrain prediction information for the excavated area and state prediction information for the working mechanism at the next moment. The third prediction model can be trained using third training samples containing the above four types of information. The label information of the third training samples is the labeled terrain prediction information and labeled state prediction information for the next moment. After training, the third prediction model will be able to accurately predict the terrain and state at the next moment based on real-time terrain detection information, target terrain information, state detection information, and historical state information.

[0128] Optionally, a fourth prediction model is pre-trained, which is a time-series neural network model. The current terrain detection information, preset target terrain information, and current state detection information are input into the pre-trained fourth prediction model, which then outputs the current terrain detection information, preset target terrain information, and current state detection information. The fourth prediction model is implemented based on a neural network structure containing time-series sequences, such as LSTM or GRU, and can memorize historical input information. Therefore, it is not necessary to input the historical state detection information of the working mechanism within the preset time period. After receiving the input current terrain detection information, preset target terrain information, and current state detection information, the fourth prediction model determines the terrain prediction information of the excavation area and the state prediction information of the working mechanism for the next moment based on the current terrain detection information, preset target terrain information, current state detection information, and the state detection information of historical input moments.

[0129] In this implementation method, the process of determining the terrain prediction information and the status prediction information of the excavation area at the next moment becomes more complex and accurate. It not only relies on the current terrain detection information and the current status detection information of the operating mechanism, but also considers the preset target terrain information and the historical status detection information of the operating mechanism in the past preset time period. That is, it not only considers the immediate environment and status, but also incorporates the target and historical data, which improves the accuracy and reliability of the prediction.

[0130] Furthermore, in addition to the current terrain detection information, the preset target terrain information, the current state detection information, and the historical state detection information of the operating mechanism within the past preset time period, the terrain prediction information of the excavation area and the state prediction information of the operating mechanism at the next moment can be determined by further combining the historical terrain detection information within the past preset time period, so as to further improve the accuracy of the prediction. The specific implementation can refer to the implementation idea of ​​the aforementioned embodiments, and will not be described in detail here.

[0131] As an optional implementation, the first strategy in step S203 is implemented based on the processing concept of dynamic change analysis, such as... Figure 3 As shown, step S203 includes steps S301-S302:

[0132] S301. Based on the terrain change between the current terrain detection information and the terrain prediction information, and the state change between the current state detection information and the state prediction information, a first pressure change is determined, wherein the first pressure change represents the predicted pressure change of the working mechanism between the current moment and the next moment.

[0133] Changes in the operating mechanism's state (such as increased speed or changes in digging depth) alter its interaction with the excavated material and the surrounding environment, leading to changes in its stress state and consequently, corresponding pressure changes. The amount of state change between the current state detection information and the state prediction information of the operating mechanism reflects the state changes that will occur between the current moment and the next moment, and can be used to assess the pressure changes that the operating mechanism may face in the next moment.

[0134] Changes in the terrain of the excavation area can alter the contact patterns and support conditions between the excavating mechanism and the excavated materials, thereby changing the pressure on the mechanism. The amount of terrain change between the current terrain detection information and the terrain prediction information of the excavation area reflects the upcoming terrain changes between the current moment and the next moment, and can be used to assess the potential pressure changes the excavating mechanism may face in the next moment.

[0135] Specifically, after obtaining the current terrain detection information and terrain prediction information of the excavation area, and the current status detection information and status prediction information of the operating mechanism, these are analyzed to determine the terrain change between the current terrain detection information and terrain prediction information of the excavation area, as well as the status change between the current status detection information and status prediction information of the operating mechanism. The above-mentioned terrain change and status change are comprehensively analyzed, taking into account the impact of the terrain change and status change on the pressure of the operating mechanism, to determine the predicted pressure change of the operating mechanism between the current time and the next time, i.e., the first pressure change.

[0136] The first pressure change reflects the possible changes in pressure on the operating mechanism from the current moment to the next moment due to changes in terrain and the state of the operating mechanism.

[0137] As an optional implementation method, after determining the amount of terrain change and state change between the current time and the next time, various methods can be used to analyze the impact of terrain change and state change on the pressure of the working mechanism, and determine the first pressure change of the working mechanism between the current time and the next time. For example, the pressure change can be estimated based on physical models or empirical formulas, or machine learning algorithms can be used to learn and predict pressure change patterns.

[0138] As an optional implementation, after determining the terrain change and state change between the current time and the next time, the terrain change and state change are input into a pre-trained pressure change determination model, and the first pressure change determination model determines the first pressure change based on the terrain change and state change.

[0139] In this implementation, a pressure change determination model is pre-trained. This model is trained using training samples that include terrain changes in the excavation area and state changes of the work mechanism. The label information for these training samples is the labeled first pressure change of the work mechanism. The trained pressure change determination model can learn the complex relationship between terrain changes, state changes, and pressure changes, and predict the first pressure change of the work mechanism based on the input data (i.e., current terrain detection information, terrain prediction information, current state detection information, and state prediction information).

[0140] Alternatively, the current terrain detection information, terrain prediction information, current state detection information, and state prediction information can be directly input into a pre-trained pressure change determination model. However, in this approach, the training samples and training method of the pressure change determination model must be modified accordingly to adapt to the new input and output. The trained pressure change determination model can determine the terrain change between the current terrain detection information and the terrain prediction information, as well as the state change between the current state detection information and the state prediction information, and predict the first pressure change of the operating mechanism based on the terrain change and state change.

[0141] S302. Based on the first pressure change and the current pressure detection information, determine the pressure prediction information of the working mechanism at the next moment.

[0142] As described above, the first pressure change reflects the change in pressure on the operating mechanism from the current moment to the next moment due to changes in terrain and the state of the operating mechanism. Therefore, after determining the first pressure change, the pressure prediction information for the operating mechanism at the next moment can be determined based on the first pressure change between the current moment and the next moment, as well as the pressure detection information at the current moment.

[0143] This implementation takes into account changes in the operating status of the working mechanism between the current moment and the next moment (such as increases or decreases in speed, changes in digging actions, etc.) and changes in terrain (such as undulations in terrain) rather than static values. It can better capture the impact of changes in operating status and terrain on the pressure on the working mechanism, thereby better adapting to the dynamic changes in the pressure of the working mechanism and providing more accurate, more adaptable to dynamic environments, more noise-resistant and more interpretable pressure prediction results.

[0144] Furthermore, as an optional implementation, steps S301 and S302 are implemented based on the processing concept of a neural network model, such as... Figure 4 As shown, step S203 includes step S401:

[0145] S401. Input the current terrain detection information, the terrain prediction information, the current state detection information, the state prediction information, and the current pressure detection information into a pre-trained first pressure prediction model. The first pressure prediction model determines a first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, and the state change between the current state detection information and the state prediction information. Based on the first pressure change and the current pressure detection information, the first pressure prediction model determines the pressure prediction information of the working mechanism at the next moment.

[0146] Specifically, in this embodiment, a first pressure prediction model is pre-trained. The first pressure prediction model is obtained by performing a second pressure prediction training on a fourth training sample. The fourth training sample includes terrain information at a first time step, terrain information at a second time step, state information at a first time step, state information at a second time step, and pressure information at a first time step, wherein the second time step is the time step following the first time step. The sample label for the fourth training sample is the pressure information at the second time step. The second pressure prediction training enables the first pressure prediction model to determine a first pressure change based on the terrain change between the terrain information at the first time step and the terrain information at the second time step, and the state change between the state information at the first time step and the state information at the second time step, and to determine the pressure prediction information at the second time step based on the first pressure change and the pressure information at the first time step.

[0147] After obtaining the current terrain detection information and terrain prediction information of the excavator, the current state detection information and state prediction information of the working mechanism, and the current pressure detection information through steps S201 and S202, this information is input into the pre-trained first pressure prediction model. The first pressure prediction model can then predict the pressure prediction information of the working mechanism at the next moment based on the amount of terrain change between the current terrain detection information and terrain prediction information of the excavation area, the amount of state change between the current state detection information and state prediction information of the working mechanism, and the current pressure detection information.

[0148] The first stress prediction model can be implemented based on various deep learning neural network structures, such as CNN (Convolutional Neural Networks), RNN (Recurrent Neural Networks), CRNN (Convolutional Recurrent Neural Networks), etc., and this application does not limit it.

[0149] Furthermore, as an optional implementation, the first pressure prediction model is a time-series neural network model. The first pressure model determines the first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, and the state change between the current state detection information and the state prediction information, including:

[0150] Based on the terrain change between the current terrain detection information and the terrain prediction information, the state change between the current state detection information and the state prediction information, and the hidden state at the previous moment, the first pressure change and the hidden state at the current moment are determined. The hidden state at the previous moment includes the model input information within a preset time period in the past and the intermediate state information corresponding to the model input information.

[0151] Specifically, the first stress prediction model can be implemented based on any type of neural network structure with inherent time series, such as RNN (Recurrent Neural Networks), LSTM (Long Short-Term Memory networks), GRU (Gated Recurrent Unit), etc., and this application does not limit it.

[0152] A time series neural network is a neural network model specifically designed for processing time series data. It combines the advantages of neural networks and time series analysis, effectively capturing time-series information and patterns in time series data and performing tasks such as prediction and classification.

[0153] In time-series neural networks, the hidden state is crucial information passed to the next time step. Calculated at the end of each time step, the hidden state contains all the important information from the historical sequence and serves as one of the inputs for the next time step (besides the external inputs for the next time step). The information in the hidden state is vital for subsequent predictions and decisions. Thus, the hidden state acts as a bridge for information transmission, enabling the network to remember and utilize past information to influence future outputs. In this implementation, the hidden state of the previous time step includes model input information from a preset time period and the intermediate state information corresponding to that model input information.

[0154] In this implementation, the first pressure prediction model can capture the temporal dependencies and dynamic changes in the input data of historical time steps, such as the temporal information and dependencies of one or more data in the excavation area terrain information, excavator operating mechanism status information, and operating mechanism pressure information, thereby assisting the current time step input information in predicting the pressure prediction information of the operating mechanism at the next moment, achieving more accurate prediction.

[0155] Preferably, in order to solve the problem of gradient vanishing or gradient exploding, the first pressure prediction model is implemented using LSTM neural networks (Long Short-Term Memory networks).

[0156] In addition to the hidden state, in LSTM (Long Short-Term Memory) neural networks, the cell state is updated at each time step based on the outputs of the forget gate and the input gate. The updated cell state not only contains relevant information from the current time step but also retains long-term dependencies from previous time steps. The updated cell state is then passed to the LSTM unit in the next time step to continue participating in subsequent computation and update processes. Unlike the hidden state, the cell state is designed to be more stable and less susceptible to short-term input fluctuations. Through fine-grained control of the forget gate, input gate, and output gate, effective preservation and updating of long-term information is achieved.

[0157] Furthermore, because the excavation cycle of construction machinery is relatively slow and highly correlated with various information during the excavation process, a weakened weight for past time moments and a feedforward mechanism for past time moments are added to the ordinary long short-term neural network structure.

[0158] Specifically, such as Figure 5 As shown in the embodiment of this application, a weight w is added to the memory cell c in the hidden layer of a conventional long short-term memory neural network. t-1 =C′ t-1 w causes the weight of important information in the early sequence recorded in memory cell c to gradually decrease as time increases.

[0159] σ represents the activation layer of the neural network, σ = F(W i A i +b i ), where W i 、b i For the neural network parameters, A i It is a state variable.

[0160] Figure 6 This is a flowchart illustrating another load prediction method provided in an embodiment of this application. Figure 6 As shown, the load prediction method provided in this application embodiment includes steps S601-S603:

[0161] S601. Obtain the current status detection information of the working mechanism, the current pressure detection information of the working mechanism, and the current terrain detection information of the excavation area. The current status detection information includes the pose detection value and / or velocity detection value of the working mechanism at the current moment.

[0162] Step S601 and Figure 2 In the illustrated embodiment, step S201 corresponds to step S601, and the specific content of step S601 can be referred to step S201, which will not be repeated here.

[0163] S602. Input the current state detection information, the current pressure detection information, and the current terrain detection information into the pre-trained second pressure prediction model to obtain the pressure prediction information of the working mechanism at the next moment.

[0164] The second pressure prediction model includes a first submodule and a second submodule. The first submodule is used to determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information and the current state detection information. The second submodule is used to determine the pressure prediction information of the working mechanism at the next moment based on the terrain prediction information, the state prediction information and the current pressure detection information.

[0165] In this embodiment, a second stress prediction model is pre-built and pre-trained.

[0166] First, a second pressure prediction model is constructed. A first submodule and a second submodule are constructed separately, and then combined according to preset rules. This ensures that in the completed second pressure prediction model, the current state detection information and current pressure detection information from the input of the second pressure prediction model serve as the input of the first submodule; the output of the first submodule (terrain prediction information of the excavation area and state prediction information of the working mechanism) and the current pressure detection information from the input of the second pressure prediction model serve as the output of the second submodule; and the output of the second submodule serves as the output of the second pressure prediction model.

[0167] Secondly, the second pressure prediction model is trained. After constructing the second pressure prediction model, the dataset consisting of the current working mechanism status information, the current excavation area terrain information, and the current working mechanism pressure information is used as the fifth training sample, along with the sample labels corresponding to the fifth training sample (labeling the working mechanism pressure information at the next moment), to train the constructed second pressure prediction model for third pressure prediction, thus obtaining the trained second pressure prediction model.

[0168] The third pressure prediction training aims to enable the second pressure prediction model to generate pressure prediction information for the working mechanism at the next moment, and to ensure that the loss function value determined based on the pressure prediction information output by the model and the pressure information of the working mechanism at the next moment meets preset requirements. The trained second pressure prediction model can determine the terrain prediction information and the state prediction information of the working mechanism at the next moment based on the current terrain detection information of the excavation area and the current state detection information of the working mechanism. Furthermore, it can determine the pressure prediction information of the working mechanism at the next moment based on the terrain prediction information, the state prediction information, and the current pressure detection information of the working mechanism.

[0169] As an optional implementation, the second pressure prediction model is a time series neural network model.

[0170] The fourth prediction model is based on neural network structures with inherent time series, such as LSTM or GRU, which can memorize historical input information. Therefore, it is not necessary to input the historical state detection information of the operating mechanism within the past preset time period.

[0171] In this implementation, optionally, the first submodule can be used to determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information, the current state detection information, the preset target terrain information, and the historical state detection information of the working mechanism in the past preset time period.

[0172] Optionally, the first submodule can be specifically used to determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information, the current state detection information, the preset target terrain information, the historical state detection information of the working mechanism in the past preset time period, and the terrain detection information of the excavation area in the past preset time period.

[0173] As an optional implementation, the second submodule can be specifically used to determine the first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, and the state change between the current state detection information and the state prediction information, and to determine the pressure prediction information of the working mechanism at the next moment based on the first pressure change and the current pressure detection information.

[0174] S603. Based on the pressure prediction information of the working mechanism at the next moment, determine the load prediction value of the working mechanism at the next moment.

[0175] Step S603 and Figure 2 In the illustrated embodiment, step S204 corresponds to step S603, and the specific content of step S603 can be referred to step S204, which will not be repeated here.

[0176] Exemplary device

[0177] Corresponding to the load prediction method described above, this application also provides a load prediction device. Figure 7 This is a schematic diagram of the structure of a load prediction device provided in an embodiment of this application, as shown below. Figure 7 As shown in the embodiment of this application, the load prediction device is applied to construction machinery, which includes a working mechanism for performing excavation operations. The device includes:

[0178] The data acquisition unit 701 is used to determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information and the current state detection information.

[0179] The first prediction unit 702 is used to determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information and the current state detection information.

[0180] The second prediction unit 703 is used to determine the pressure prediction information of the working mechanism at the next moment based on the terrain prediction information, the state prediction information and the current pressure detection information, wherein the pressure prediction information is used to determine the load of the engineering machinery.

[0181] The load prediction unit 704 is used to determine the load prediction value of the working mechanism at the next moment based on the pressure prediction information of the working mechanism at the next moment.

[0182] The load prediction device provided in this application embodiment, on the one hand, can continuously and accurately predict the pressure prediction information of the working mechanism at the next moment by acquiring the current status detection information of the working mechanism, the current pressure detection information of the working mechanism, and the current terrain detection information of the excavation area in real time, thereby predicting the load situation of the working mechanism at the next moment, meeting the real-time and accuracy requirements of excavating machinery for load prediction; on the other hand, it does not require the establishment of complex soil and material models, reducing the complexity of calculation and improving the real-time performance of prediction. Moreover, it does not depend on specific soil and material parameters, reducing prediction errors caused by changes in soil or material types, which is in line with the complex and ever-changing actual working conditions of excavating machinery and has wider applicability.

[0183] Optionally, the first prediction unit 702 can be specifically used for:

[0184] Based on the current terrain detection information, the preset target terrain information, the current status detection information, and the historical status detection information of the working mechanism within the past preset time period, the terrain prediction information of the excavation area and the status prediction information of the working mechanism at the next moment are determined.

[0185] Optionally, the second prediction unit 703 can be specifically used for:

[0186] Based on the amount of terrain change between the current terrain detection information and the terrain prediction information, and the amount of state change between the current state detection information and the state prediction information, a first pressure change is determined, wherein the first pressure change represents the predicted pressure change of the working mechanism between the current moment and the next moment.

[0187] Based on the first pressure change and the current pressure detection information, the pressure prediction information of the working mechanism at the next moment is determined.

[0188] Optionally, the second prediction unit 703 can be specifically used for:

[0189] The current terrain detection information, the terrain prediction information, the current state detection information, the state prediction information, and the current pressure detection information are input into a pre-trained first pressure prediction model. The first pressure prediction model determines a first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, and the state change between the current state detection information and the state prediction information. Based on the first pressure change and the current pressure detection information, the first pressure prediction model determines the pressure prediction information of the working mechanism at the next moment.

[0190] Optionally, the first pressure prediction model is a time series neural network model, and the step of determining the first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, and the state change between the current state detection information and the state prediction information, includes:

[0191] Based on the terrain change between the current terrain detection information and the terrain prediction information, the state change between the current state detection information and the state prediction information, and the hidden state at the previous moment, the first pressure change and the hidden state at the current moment are determined. The hidden state at the previous moment includes the model input information within a preset time period in the past and the intermediate state information corresponding to the model input information.

[0192] Optionally, the first prediction unit 702 and the second prediction unit 703 can be combined into one unit, which can be used for:

[0193] The current state detection information, the current pressure detection information, and the current terrain detection information are input into a pre-trained second pressure prediction model to obtain the pressure prediction information of the working mechanism at the next moment. The second pressure prediction model is a time series neural network model, which includes a first sub-module and a second sub-module.

[0194] The first submodule is used to determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism in the next moment based on the current terrain detection information, the current state detection information, the preset target terrain information, and the historical state detection information within the past preset time period in the hidden state of the previous moment.

[0195] The second submodule is used to determine the first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, the state change between the current state detection information and the state prediction information, and the hidden state at the previous moment, and to determine the pressure prediction information of the working mechanism at the next moment based on the first pressure change and the current pressure detection information. The hidden state at the previous moment includes the model input information within a preset time period in the past and the intermediate state information corresponding to the model input information.

[0196] The load prediction device provided in this embodiment belongs to the same concept as the load prediction method provided in the above embodiments of this application. It can execute the load prediction method provided in any of the above embodiments of this application and has the corresponding functional modules and beneficial effects for executing the load prediction method. Technical details not described in detail in this embodiment can be found in the specific processing content of the load prediction method provided in the above embodiments of this application, and will not be repeated here.

[0197] Exemplary electronic devices

[0198] Another embodiment of this application also proposes a load prediction device, see [link to relevant documentation] Figure 8 As shown, the device includes:

[0199] Memory 200 and processor 210;

[0200] The memory 200 is connected to the processor 210 and is used to store programs;

[0201] The processor 210 is configured to implement the load prediction method disclosed in any of the above embodiments by running the program stored in the memory 200.

[0202] Specifically, the aforementioned load prediction device may also include: a bus, a communication interface 220, an input device 230, and an output device 240.

[0203] The processor 210, memory 200, communication interface 220, input device 230, and output device 240 are interconnected via a bus. Among them:

[0204] A bus can include a pathway for transmitting information between various components of a computer system.

[0205] The processor 210 can be a general-purpose processor, such as a general-purpose central processing unit (CPU), a microprocessor, etc., or an application-specific integrated circuit (ASIC), or one or more integrated circuits used to control the execution of the program of the present invention. It can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0206] Processor 210 may include a main processor, as well as a baseband chip, modem, etc.

[0207] The memory 200 stores a program that executes the technical solution of this invention, and may also store an operating system and other key business functions. Specifically, the program may include program code, which includes computer operation instructions. More specifically, the memory 200 may include read-only memory (ROM), other types of static storage devices capable of storing static information and instructions, random access memory (RAM), other types of dynamic storage devices capable of storing information and instructions, disk storage, flash memory, etc.

[0208] Input device 230 may include a device for receiving user input data and information, such as a keyboard, mouse, camera, scanner, light pen, voice input device, touch screen, pedometer, or gravity sensor.

[0209] Output device 240 may include devices that allow information to be output to a user, such as a display screen, printer, speaker, etc.

[0210] The communication interface 220 may include a device that uses any transceiver to communicate with other devices or communication networks, such as Ethernet, Radio Access Network (RAN), Wireless Local Area Network (WLAN), etc.

[0211] The processor 210 executes the program stored in the memory 200 and calls other devices, which can be used to implement the various steps of any of the load prediction methods provided in the above embodiments of this application.

[0212] Exemplary computer program products and storage media

[0213] In addition to the methods and apparatus described above, embodiments of this application may also be computer program products, which include computer program instructions that, when executed by a processor, cause the processor to perform the steps in the load prediction methods according to various embodiments of this application as described in the "Exemplary Methods" section of this specification.

[0214] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this application. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0215] Furthermore, embodiments of this application may also be storage media storing a computer program, which is executed by a processor in the load prediction method according to various embodiments of this application described in the "Exemplary Methods" section above. Specifically, the following steps can be implemented:

[0216] S201. Obtain the current status detection information of the working mechanism, the current pressure detection information of the working mechanism, and the current terrain detection information of the excavation area. The current status detection information includes the pose detection value and / or velocity detection value of the working mechanism at the current moment.

[0217] S202. Based on the current terrain detection information and the current state detection information, determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment.

[0218] S203. Based on the terrain prediction information, the state prediction information, and the current pressure detection information, determine the pressure prediction information of the working mechanism at the next moment.

[0219] S204. Based on the pressure prediction information of the working mechanism at the next moment, determine the load prediction value of the working mechanism at the next moment.

[0220] For the foregoing method embodiments, in order to simplify the description, they are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to this application, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0221] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For apparatus embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0222] The steps in the methods of the various embodiments of this application can be adjusted, merged, or deleted in order according to actual needs, and the technical features described in each embodiment can be replaced or combined.

[0223] The modules and sub-modules in the apparatus and terminal in the various embodiments of this application can be merged, divided, and deleted according to actual needs.

[0224] It should be understood that the disclosed terminals, devices, and methods can be implemented in other ways, given the several embodiments provided in this application. For example, the terminal embodiments described above are merely illustrative. For instance, the division of modules or sub-modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple sub-modules or modules may be combined or integrated into another module, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or modules, and may be electrical, mechanical, or other forms.

[0225] The modules or submodules described as separate components may or may not be physically separate. The components that constitute a module or submodule may or may not be physical modules or submodules; that is, they may be located in one place or distributed across multiple network modules or submodules. Some or all of the modules or submodules can be selected to achieve the purpose of this embodiment's solution, depending on actual needs.

[0226] Furthermore, the functional modules or sub-modules in the various embodiments of this application can be integrated into one processing module, or each module or sub-module can exist physically separately, or two or more modules or sub-modules can be integrated into one module. The integrated modules or sub-modules described above can be implemented in hardware or in the form of software functional modules or sub-modules.

[0227] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0228] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software unit executed by a processor, or a combination of both. The software unit can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0229] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0230] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A load forecasting method, characterized in that, Applied to construction machinery, the construction machinery including a working mechanism for performing excavation operations, the method includes: The current status detection information of the working mechanism, the current pressure detection information of the working mechanism, and the current terrain detection information of the excavation area are obtained. The current status detection information includes the pose detection value and / or velocity detection value of the working mechanism at the current moment. Based on the current terrain detection information and the current status detection information, determine the terrain prediction information of the excavation area and the status prediction information of the working mechanism at the next moment; Based on the terrain prediction information, the state prediction information, and the current pressure detection information, the pressure prediction information of the working mechanism at the next moment is determined; Based on the amount of terrain change between the current terrain detection information and the terrain prediction information, and the amount of state change between the current state detection information and the state prediction information, a first pressure change is determined, wherein the first pressure change represents the predicted pressure change of the working mechanism between the current moment and the next moment. Based on the first pressure change and the current pressure detection information, the pressure prediction information of the working mechanism at the next moment is determined; The current terrain detection information, the terrain prediction information, the current state detection information, the state prediction information, and the current pressure detection information are input into a pre-trained first pressure prediction model. The first pressure prediction model determines a first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, and the state change between the current state detection information and the state prediction information. Based on the first pressure change and the current pressure detection information, the first pressure prediction model determines the pressure prediction information of the working mechanism at the next moment. Based on the pressure prediction information of the operating mechanism at the next moment, determine the load prediction value of the operating mechanism at the next moment.

2. The method according to claim 1, characterized in that, The step of determining the predicted terrain information of the excavation area and the predicted state information of the working mechanism at the next moment based on the current terrain detection information and the current state detection information includes: Based on the current terrain detection information, the preset target terrain information, the current status detection information, and the historical status detection information of the working mechanism within the past preset time period, the terrain prediction information of the excavation area and the status prediction information of the working mechanism at the next moment are determined.

3. The method according to claim 1, characterized in that, The first pressure prediction model is a time series neural network model. The step of determining the first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, and the state change between the current state detection information and the state prediction information, includes: Based on the terrain change between the current terrain detection information and the terrain prediction information, the state change between the current state detection information and the state prediction information, and the hidden state at the previous moment, the first pressure change and the hidden state at the current moment are determined. The hidden state at the previous moment includes the model input information within a preset time period in the past and the intermediate state information corresponding to the model input information.

4. The method according to claim 1, characterized in that, The step of determining the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information and the current state detection information, and determining the pressure prediction information of the working mechanism at the next moment based on the terrain prediction information, the state prediction information, and the current pressure detection information, includes: The current state detection information, the current pressure detection information, and the current terrain detection information are input into a pre-trained second pressure prediction model to obtain the pressure prediction information of the working mechanism at the next moment. The second pressure prediction model is a time series neural network model, which includes a first sub-module and a second sub-module. The first submodule is used to determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism in the next moment based on the current terrain detection information, the current state detection information, the preset target terrain information, and the historical state detection information within the past preset time period in the hidden state of the previous moment. The second submodule is used to determine the first pressure change based on the terrain change between the current terrain detection information and the terrain prediction information, the state change between the current state detection information and the state prediction information, and the hidden state at the previous moment, and to determine the pressure prediction information of the working mechanism at the next moment based on the first pressure change and the current pressure detection information. The hidden state at the previous moment includes the model input information within a preset time period in the past and the intermediate state information corresponding to the model input information.

5. A load prediction device, characterized in that, The load prediction method according to claim 1 is applied to construction machinery, the construction machinery including a working mechanism for performing excavation operations, the device comprising: The data acquisition unit is used to acquire the current status detection information of the working mechanism, the current pressure detection information of the working mechanism, and the current terrain detection information of the excavation area. The current status detection information includes the pose detection value and / or velocity detection value of the working mechanism at the current moment. The first prediction unit is used to determine the terrain prediction information of the excavation area and the state prediction information of the working mechanism at the next moment based on the current terrain detection information and the current state detection information. The second prediction unit is used to determine the pressure prediction information of the working mechanism at the next moment based on the terrain prediction information, the state prediction information and the current pressure detection information; The load prediction unit is used to determine the load prediction value of the working mechanism at the next moment based on the pressure prediction information of the working mechanism at the next moment.

6. A computer program product, characterized in that, It includes computer program instructions that, when executed by a processor, cause the processor to implement the load prediction method as described in any one of claims 1-4.

7. A load prediction device, characterized in that, Including memory and processor; The memory is connected to the processor and is used to store programs; The processor is used to implement the load prediction method as described in any one of claims 1-4 by running a program in the memory.

8. An engineering machinery, characterized in that, include: The operating mechanism, the detection equipment, and the load prediction device as described in claim 7, wherein, The detection equipment includes a first detection device, a second detection device, and a third detection device, and the first detection device, the second detection device, and the third detection device are respectively communicatively connected to the load prediction device. The working mechanism is used to perform excavation operations; The first detection device is used to collect the current status detection value of the working mechanism in real time; The second detection device is used to collect the current pressure detection value of the operating mechanism in real time; The third detection device is used to collect terrain information of the area to be excavated in real time.

9. The engineering machinery according to claim 8, characterized in that, The construction machinery is an excavator; The working mechanism includes a boom, a stick, and a bucket; The first detection equipment includes boom condition detection equipment, stick condition detection equipment, and bucket condition detection equipment; The second testing equipment includes boom pressure testing equipment, stick pressure testing equipment, and bucket pressure testing equipment.

Images