An excavator autonomous work control method and system

Abstract

The present application relates to a kind of excavator autonomous operation control method and system, belong to engineering machinery control technical field.The present application provides a kind of excavator autonomous operation control method and system, after the initialization operation preparation of excavator is completed, the standard operation model obtained by the pre-regression prediction corresponding to target operation action is called, the autonomous operation control of excavator is carried out according to the order of working condition step, the standard operation model includes the autonomous operation model of each working condition and the autonomous switching model of the switching between adjacent working conditions;The present application carries out whole working condition operation by including the autonomous operation model of each working condition and the autonomous switching model of the switching between adjacent working conditions in standard operation model, and according to the pressure change curve of the pressure of main pump under current working condition and the working time of each component as independent variable, constructs working condition adaptive switching function to make each working condition form clear distinction, to ensure that the whole autonomous operation process can be carried out smoothly according to the order of working condition step.

Description

Technical Field

[0001] This invention relates to an autonomous operation control method and system for excavators, belonging to the field of engineering machinery control technology. Background Technology

[0002] Excavators are essential pieces of equipment widely used in engineering construction, capable of performing various construction operations such as digging, leveling, loading and unloading, slope trimming, breaking, scraping, lifting, and towing. Traditional excavators primarily rely on operators manually operating control levers to drive the boom, stick, bucket, and slewing actuators in single or combined movements to complete digging, leveling, loading and unloading, slope trimming, and breaking operations. The efficiency and quality of these operations are significantly affected by the operator's skill level. Although some excavators have been equipped with auxiliary functions, they still suffer from low levels of automation, complex operation, limited application scenarios, and compromised safety during construction.

[0003] With the development of electrification and intelligentization of excavators, unmanned excavators are gradually playing a role in engineering construction. However, existing unmanned excavator technology still suffers from unclear distinctions between working conditions and chaotic progress, leading to frequent overworking or misoperation during operation, and even causing the excavator to tilt at a large angle and overturn, posing significant safety hazards.

[0004] Therefore, there is an urgent need for a technical solution that can reliably achieve single-condition operation and smooth transition between operating conditions. Summary of the Invention

[0005] The purpose of this invention is to provide an autonomous operation control method and system for excavators. The system executes full-condition operation through a standard operation model that includes autonomous operation models for various working conditions and autonomous switching models for switching between adjacent working conditions. Furthermore, it constructs a working condition adaptation switching function based on the pressure change curve of the main pump pressure under the current working condition and the working time of each component as independent variables to clearly distinguish between various working conditions, thereby ensuring that the entire autonomous operation process can proceed smoothly according to the working condition steps.

[0006] In a first aspect, the present invention provides an autonomous operation control method for an excavator, comprising: Based on the environmental information of the excavator's current operating scenario, determine the target operation action that needs to be performed; The parameter ranges of the working parameters of each working component of the excavator are initialized according to the target operation action, and then the excavator is prepared for initialization operation. After the excavator completes its initialization preparation, it calls the standard operation model obtained from the pre-regression prediction corresponding to the target operation action and performs autonomous operation control of the excavator according to the working condition steps. The standard operation model includes autonomous operation models for each working condition and autonomous switching models for switching between adjacent working conditions. Within each working condition, the corresponding autonomous operation model is invoked. The autonomous operation model infers and outputs the operation control curve of each working component under the current working condition based on the real-time values ​​of the working parameters of each working component and the corresponding parameter range, so as to control the excavator to perform autonomous operation. When the working time of each working condition reaches the set time, it is determined whether to switch to the next working condition based on the working condition adaptation switching function obtained from pre-supervised training. The independent variables of the working condition adaptation switching function are the pressure change curve of the main pump pressure under the current working condition and the working time of each component. If the judgment result is yes, the corresponding autonomous switching model is called to output the working condition switching control curve of each working component. After the working condition switching control curve ends, the autonomous operation model of the next working condition is called to perform autonomous operation. If the judgment result is no, the real-time values ​​of the working parameters of each working component are fine-tuned within the parameter range until the judgment result of the working condition adaptation switching function is yes.

[0007] Optionally, the excavator's initialization preparation includes: Set the initial working parameters of each working component when the excavator's autonomous operation function is activated; Before starting autonomous operation, the excavator's boom and bucket are swung out to their extreme positions, the boom is adjusted to make the bucket touch the ground, and the swing remains unchanged to keep the bucket facing forward, completing the posture preparation before autonomous operation, which serves as the prior information for calling the standard operation model to execute actions.

[0008] Optionally, the method for constructing the standard operating procedure model includes: Acquire full-condition manual control data for the target operation; The full-condition manual operation data is input into the two-layer time-series operation model for regression prediction to obtain the standard operation model.

[0009] Optionally, inputting the full-condition manual operation data into a two-layer time-series operation model for regression prediction to obtain a standard operation model includes: Extract the single-condition manual control data for each working condition from the full-condition manual control data; Extract the dual-condition manual control data for each pair of adjacent conditions from the full-condition manual control data; For each working condition, the corresponding single-working-condition manual control data is input into the bottom time series model of the two-layer time series operation model. The bottom time series model uses short time window learning to capture the fast dynamic features of the single-working-condition manual control data in order to obtain the autonomous operation model of the current working condition. For each pair of adjacent working conditions, the corresponding manual control data of the two working conditions is input into the high-level time series model of the two-layer time series operation model. The high-level time series model uses a long time window to capture and process the long-distance time series dependency relationship of the manual control data of the two working conditions in order to obtain the autonomous switching model between the pair of adjacent working conditions.

[0010] Optionally, the standard operating model utilizes knowledge distillation technology for lightweight deployment in low-computing-power scenarios.

[0011] Optionally, the working components include a bucket, a stick, a boom, a slewing mechanism, a left travel mechanism, and a right travel mechanism.

[0012] Optionally, the target operation is a trenching operation, and the trenching operation includes digging, lifting and rotating, unloading, resetting and reversing conditions.

[0013] Optionally, during the process of autonomously controlling the excavator according to the sequence of working conditions, real-time safety control is also required. Methods for real-time safety control include: Real-time information on obstacles near the excavator is acquired. If an obstacle is within the operating range of at least one working component, an emergency stop is triggered to control all working components to stop operating.

[0014] Optionally, within the parameter range, the real-time values ​​of the working parameters of each working component are fine-tuned until the judgment result of the working condition adaptation switching function is yes. If the main pump pressure reaches the preset system pressure, the judgment result of the working condition adaptation switching function is directly determined to be yes and the corresponding autonomous switching model is called.

[0015] In a second aspect, the present invention provides an autonomous operation control system for an excavator, comprising: The determination module is used to determine the target operation action to be performed based on the environmental information of the excavator's current working scenario. The initialization module is used to initialize the parameter range of the working parameters of each working component of the excavator according to the target operation action after the excavator has completed the initialization operation preparation. After the initialization setting is completed, the standard operation model corresponding to the target operation action obtained by pre-regression prediction is called to perform autonomous operation control of the excavator according to the working condition steps. The standard operation model includes autonomous operation models for each working condition and autonomous switching models for switching between adjacent working conditions. The autonomous operation module is used to call the corresponding autonomous operation model for each working condition. The autonomous operation model, based on the real-time values ​​and corresponding parameter ranges of the working parameters of each working component, infers and outputs the operation control curves of each working component under the current working condition to control the excavator to perform autonomous operation. When the operation time for each working condition reaches the set time, it determines whether to switch to the next working condition based on the pre-trained working condition adaptation switching function. The independent variables of the working condition adaptation switching function are the pressure change curve of the main pump pressure and the working time of each component under the current working condition. If the determination result is yes, the corresponding autonomous switching model is called to output the working condition switching control curves of each working component. After the working condition switching control curves are completed, the autonomous operation model for the next working condition is called to perform autonomous operation. If the determination result is no, the real-time values ​​of the working parameters of each working component are fine-tuned within the parameter range until the determination result of the working condition adaptation switching function is yes.

[0016] Compared with existing technologies, this invention provides an autonomous operation control method and system for excavators, which has the following beneficial effects: Based on the environmental information of the excavator's current operating scenario, this invention determines the target operation action to be performed; based on the target operation action, it initializes the parameter range of the working parameters of each working component of the excavator, and then performs initialization preparation for the excavator; after the excavator's initialization preparation is completed, it calls the standard operation model corresponding to the target operation action obtained through pre-regression prediction, and performs autonomous operation control of the excavator according to the working condition steps. The standard operation model includes autonomous operation models for each working condition and autonomous switching models for switching between adjacent working conditions; wherein, the standard operation model is obtained by regression prediction based on the input of full-condition manual control data into a two-layer time-series operation model.

[0017] Within each working condition, the corresponding autonomous operation model is invoked. Based on the parameter range of each working component's operating parameters, the autonomous operation model outputs the operation control curves of each working component under the current working condition to control the excavator to perform autonomous operation. When the operation time for each working condition reaches the set time, a working condition adaptation switching function obtained through pre-supervised training is used to determine whether to switch to the next working condition. The independent variables of the working condition adaptation switching function are the pressure change curve of the main pump pressure and the working time of each component under the current working condition. If the determination result is yes, the corresponding autonomous switching model is invoked to output the working condition switching control curves of each working component. After the working condition switching control curves are completed, the autonomous operation model for the next working condition is invoked to perform autonomous operation. If the determination result is no, the real-time values ​​of the operating parameters of each working component are fine-tuned within the parameter range until the working condition adaptation switching function's determination result is yes. This invention executes full-condition operation through a standard operation model that includes autonomous operation models for each working condition and autonomous switching models for switching between adjacent working conditions. Furthermore, it constructs a working condition adaptive switching function based on the pressure change curve of the main pump under the current working condition and the working time of each component as independent variables to clearly distinguish and smoothly switch between different working conditions, thereby ensuring that the entire autonomous operation process can proceed smoothly according to the working condition steps. It is worth noting that the standard operation model in this invention utilizes knowledge distillation technology for lightweight deployment on the excavator, implemented on a hardware platform based on a low-computing-power controller, enabling autonomous operation in resource-constrained application scenarios. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the excavator autonomous operation control method of the present invention; Figure 2 This is a diagram of the architecture of the excavator to which the excavator autonomous operation control method of the present invention is applied. Detailed Implementation

[0019] It should be noted that:

[0020] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations thereof. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0021] The term "and / or" simply describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0022] Combination Figure 1 This embodiment provides an autonomous operation control method for excavators, which includes: Step S1: Based on the environmental information of the excavator's current operating scenario, determine the target operation action to be performed. For example, the target operation action could be standard trenching, slope repair, loading, dumping, or crushing operations. The corresponding target operation action can be determined based on the environmental information of the operating scenario. In actual operation, the user can select the target operation action through the handle, buttons, and instrument panel. During autonomous control, the instrument panel can interact with the user in real time, facilitating timely intervention in case of operational abnormalities. This embodiment develops a setting switch for the corresponding operation on the instrument panel, allowing the user to choose whether to activate automatic operations, such as automatic leveling or automatic trenching, achieving flexible switching of operating modes and improving the convenience and flexibility of operation.

[0023] Step S2: Initialize the parameter range of the working parameters of each working component of the excavator according to the target operation action, and then prepare the excavator for initialization operation. As an example, combined Figure 2 The working components of an excavator include a bucket, stick, boom, swing, left travel, and right travel. Specifically, the parameter range of each working component can generally be set according to the factory parameters. However, due to the diverse operating scenarios of excavators, such as in some narrow environments, it is necessary to limit the parameter range of the working parameters of the bucket, stick, boom, swing, left travel, and right travel to ensure operational safety. The working parameters are usually the basic parameters of the hydraulic drive. The main pump is connected to the drive oil port of each working component through an electromagnetic proportional valve. A pressure sensor for collecting the pressure of the main pump is installed at the oil outlet of the main pump. An overflow valve for setting the system pressure is connected to the oil outlet of the main pump. The electromagnetic proportional valves and pressure sensors of each working component are all communicatively connected to the excavator's vehicle controller. The vehicle controller is also communicatively connected to a monitor and a safety handle for controlling the pilot pump. The main body for executing the steps in this embodiment is the autonomous communication module that is communicatively connected to the vehicle controller.

[0024] Specifically, the initialization preparation of the excavator includes: setting the initial setting parameters for enabling the excavator's autonomous operation function; before starting autonomous operation, the excavator's stick and bucket are swung out to their extreme positions, the boom is adjusted to make the bucket contact the ground, and the slewing mechanism remains unchanged to keep the bucket in front, completing the posture preparation before autonomous operation, which serves as the prior information for calling the standard operation model to execute actions.

[0025] Step S3: After the excavator initialization operation preparation is completed, the standard operation model corresponding to the target operation action obtained by pre-regression prediction is called, and the excavator's autonomous operation control is performed according to the working condition step sequence. The standard operation model includes autonomous operation models for each working condition and autonomous switching models for switching between adjacent working conditions. The construction method of the standard operation model includes: obtaining full-condition manual control data of the target operation action; inputting the full-condition manual control data into the two-layer time-series operation model for regression prediction to obtain the standard operation model.

[0026] The process of inputting the full-condition manual operation data into a two-layer time-series operation model for regression prediction to obtain a standard operation model includes: extracting single-condition manual operation data for each operation condition from the full-condition manual operation data; extracting dual-condition manual operation data for each pair of adjacent operation conditions from the full-condition manual operation data; for each operation condition, inputting the corresponding single-condition manual operation data into the bottom-level time-series model of the two-layer time-series operation model, where the bottom-level time-series model uses a short window to learn and capture the fast dynamic features of the single-condition manual operation data to obtain an autonomous operation model for the current operation condition; for each pair of adjacent operation conditions, inputting the corresponding dual-condition manual operation data into the top-level time-series model of the two-layer time-series operation model, where the top-level time-series model uses a long window to capture and process the long-distance temporal dependencies of the dual-condition manual operation data to obtain an autonomous switching model for switching between the pair of adjacent operation conditions; the short window duration of the bottom-level time-series model is 200ms-1000ms, and the long window duration of the top-level time-series model is 1s-4s. In order to obtain a smoother control curve that is closer to reality, this embodiment uses a low-level timing model with a short window and a high-level timing model with a long window to process single-condition manual control data and dual-condition manual control data, respectively, so as to obtain autonomous operation models for each condition and autonomous switching models for switching between adjacent conditions.

[0027] Furthermore, the standard operating procedure model in this embodiment utilizes knowledge distillation technology to achieve lightweight and deployable operation. In the specific operation process, full-condition manual control data is input into the teacher model, and the generalized additive model (GAM) is used as the student model to perform knowledge transfer under model compression, thereby establishing a lightweight and deployable standard operating procedure model.

[0028] The total loss function for training the standard operation model is: ; in, It is a supervised loss based on the real label. It is the distillation loss between teacher output and student output. To balance hyperparameters, This represents the prediction results from the teacher model.

[0029] Step S4: Within each working condition, the corresponding autonomous operation model is invoked. Based on the real-time values ​​and corresponding parameter ranges of the working parameters of each working component, the autonomous operation model infers and outputs the operation control curves of each working component under the current working condition to control the excavator to perform autonomous operation. In this embodiment, the working components include the bucket, stick, boom, swing, left travel, and right travel. The operation control curves of each working component are represented as follows: Where i can be 1-6, representing bucket, stick, boom, swing, left travel, and right travel respectively. The autonomous operation model calculates the control signals for executing the actions of each working component at the current moment. Real-time The output is sent to the excavator's main controller; the main controller will... As the input signal for motion control, it then realizes the autonomous movement of each working component based on the control logic of its internal hydraulic, power, electrical and other subsystems.

[0030] Step S5: When the operating time for each working condition reaches the set time, determine whether to switch to the next working condition based on the working condition adaptation switching function obtained from pre-supervised training. The independent variables of the working condition adaptation switching function are the pressure change curve of the main pump pressure under the current working condition and the working time of each component. If the determination result is yes, the corresponding autonomous switching model is called to output the working condition switching control curve of each working component. After the working condition switching control curve is completed, the autonomous operation model for the next working condition is called to perform autonomous operation. If the determination result is no, the real-time values ​​of the working parameters of each working component are fine-tuned within the parameter range until the determination result of the working condition adaptation switching function is yes. If the main pump pressure reaches the system pressure set by the overflow valve, the determination result of the working condition adaptation switching function is directly determined to be yes.

[0031] The excavator's autonomous operation control module adapts to the switching function based on working conditions. The monitoring information determines whether to switch to the next working condition.

[0032] in, and The pressure of main pump one and main pump two, t is the time variable for the execution of the current working condition action. The working time for the bucket, stick, boom, swing, left travel, and right travel respectively under the current working conditions is explained in conjunction with the actual operation process.

[0033] This embodiment explains why the operating condition is determined based on the pressure change curve of the main pump and the working time of each component. Taking trenching as an example, the operating conditions of trenching include digging, lifting and slewing, unloading, resetting, and reversing. The working time of each component directly reflects the posture and progress of the operation, so it is used as the basis for the operating condition adaptation switching function. However, the working time alone cannot determine whether the excavator is performing effective load work. Therefore, the operating condition is also determined by the main pump pressure, which directly reflects load changes. The load characteristics of digging are large loads that gradually increase from minimum to maximum. The load characteristics of lifting and slewing are a certain load with relatively stable load information. The load characteristics of unloading are that the load rapidly decreases from maximum to minimum. The load of resetting is the minimum and remains relatively stable. The load of reversing is small and can be determined by the walking status. Therefore, the operating condition can be determined based on the pressure change curve of the main pump and the working time of each component.

[0034] During autonomous operation, if the main pump pressure becomes clearly abnormal, the safety interlock unit needs to intervene to stop the operation of each working component, requiring manual intervention for adjustment. In addition to using the working condition adaptation switching function as a judgment basis, the environmental perception module also needs to provide safety warnings during actual operation. Once the environmental perception module determines that there is a working hazard (excavator overload) or a safety hazard (personnel or equipment within the range of operation of the working components), a risk emergency stop is required to halt the automatic operation of each working component.

[0035] Another specific embodiment provides an autonomous operation control system for an excavator, which includes: The determination module is used to determine the target operation action to be performed based on the environmental information of the excavator's current working scenario. The initialization module is used to initialize the parameter range of the working parameters of each working component of the excavator according to the target operation action after the excavator has completed the initialization operation preparation. After the initialization setting is completed, the standard operation model corresponding to the target operation action obtained by pre-regression prediction is called to perform autonomous operation control of the excavator according to the working condition steps. The standard operation model includes autonomous operation models for each working condition and autonomous switching models for switching between adjacent working conditions. The autonomous operation module is used to call the corresponding autonomous operation model for each working condition. The autonomous operation model outputs the operation control curve of each working component under the current working condition based on the parameter range of the working parameters of each working component, and performs autonomous operation. When the operation time of each working condition reaches the set time, it determines whether to switch to the next working condition based on the working condition adaptation switching function obtained through pre-supervised training. The independent variables of the working condition adaptation switching function are the pressure change curve of the main pump pressure and the working time of each component under the current working condition. If the judgment result is yes, the corresponding autonomous switching model is called to output the working condition switching control curve of each working component. After the working condition switching control curve ends, the autonomous operation model of the next working condition is called to perform autonomous operation. If the judgment result is no, the real-time values ​​of the working parameters of each working component are fine-tuned within the parameter range until the judgment result of the working condition adaptation switching function is yes.

[0036] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0037] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0038] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0039] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0040] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A method for autonomous operation control of an excavator, characterized in that, include: Based on the environmental information of the excavator's current operating scenario, determine the target operation action that needs to be performed; The parameter ranges of the working parameters of each working component of the excavator are initialized according to the target operation action, and then the excavator is prepared for initialization operation. After the excavator completes its initialization preparation, it calls the standard operation model obtained from the pre-regression prediction corresponding to the target operation action and performs autonomous operation control of the excavator according to the working condition steps. The standard operation model includes autonomous operation models for each working condition and autonomous switching models for switching between adjacent working conditions. Within each working condition, the corresponding autonomous operation model is invoked. The autonomous operation model infers and outputs the operation control curve of each working component under the current working condition based on the real-time values ​​of the working parameters of each working component and the corresponding parameter range, so as to control the excavator to perform autonomous operation. When the working time of each working condition reaches the set time, it is determined whether to switch to the next working condition based on the working condition adaptation switching function obtained from pre-supervised training. The independent variables of the working condition adaptation switching function are the pressure change curve of the main pump pressure under the current working condition and the working time of each component. If the judgment result is yes, the corresponding autonomous switching model is called to output the working condition switching control curve of each working component. After the working condition switching control curve ends, the autonomous operation model of the next working condition is called to perform autonomous operation. If the judgment result is no, the real-time values ​​of the working parameters of each working component are fine-tuned within the parameter range until the judgment result of the working condition adaptation switching function is yes.

2. The excavator autonomous operation control method according to claim 1, characterized in that, The initial setup for the excavator includes: Set the initial working parameters of each working component when the excavator's autonomous operation function is activated; Before starting autonomous operation, the excavator's boom and bucket are swung out to their extreme positions, the boom is adjusted to make the bucket touch the ground, and the swing remains unchanged to keep the bucket facing forward, completing the posture preparation before autonomous operation, which serves as the prior information for calling the standard operation model to execute actions.

3. The method for autonomous operation control of an excavator according to claim 1, characterized in that, The method for constructing the standard operating procedure model includes: Acquire full-condition manual control data for the target operation; The full-condition manual operation data is input into the two-layer time-series operation model for regression prediction to obtain the standard operation model.

4. The excavator autonomous operation control method according to claim 3, characterized in that, The standard operation model is obtained by inputting the full-condition manual operation data into a two-layer time-series operation model for regression prediction, including: Extract the single-condition manual control data for each working condition from the full-condition manual control data; Extract the dual-condition manual control data for each pair of adjacent conditions from the full-condition manual control data; For each working condition, the corresponding single-working-condition manual control data is input into the bottom time series model of the two-layer time series operation model. The bottom time series model uses short time window learning to capture the fast dynamic features of the single-working-condition manual control data in order to obtain the autonomous operation model of the current working condition. For each pair of adjacent working conditions, the corresponding manual control data of the two working conditions is input into the high-level time series model of the two-layer time series operation model. The high-level time series model uses a long time window to capture and process the long-distance time series dependency relationship of the manual control data of the two working conditions in order to obtain the autonomous switching model between the pair of adjacent working conditions.

5. The method for autonomous operation control of an excavator according to claim 1, characterized in that, The standard operating model utilizes knowledge distillation technology for lightweight deployment in low-computing-power scenarios.

6. The method for autonomous operation control of an excavator according to claim 1, characterized in that, The working components include a bucket, a stick, a boom, a slewing mechanism, a left travel mechanism, and a right travel mechanism.

7. The method for autonomous operation control of an excavator according to claim 1, characterized in that, The target operation is trenching, and the trenching operation includes digging, lifting and rotating, unloading, resetting and reversing.

8. The method for autonomous operation control of an excavator according to claim 1, characterized in that, The process of autonomously controlling the excavator according to the working condition steps also requires real-time safety control. Real-time safety control methods include: Real-time information on obstacles near the excavator is acquired. If an obstacle is within the operating range of at least one working component, an emergency stop is triggered to control all working components to stop operating.

9. The method for autonomous operation control of an excavator according to claim 1, characterized in that, Within the parameter range, the real-time values ​​of the working parameters of each working component are fine-tuned until the judgment result of the working condition adaptation switching function is yes. If the main pump pressure reaches the preset system pressure, the judgment result of the working condition adaptation switching function is directly determined to be yes and the corresponding autonomous switching model is called.

10. An autonomous operation control system for an excavator, characterized in that, include: The determination module is used to determine the target operation action to be performed based on the environmental information of the excavator's current working scenario. The initialization module is used to initialize the parameter range of the working parameters of each working component of the excavator according to the target operation action after the excavator has completed the initialization operation preparation. After the initialization setting is completed, the standard operation model corresponding to the target operation action obtained by pre-regression prediction is called to perform autonomous operation control of the excavator according to the working condition steps. The standard operation model includes autonomous operation models for each working condition and autonomous switching models for switching between adjacent working conditions. The autonomous operation module is used to call the corresponding autonomous operation model for each working condition. The autonomous operation model, based on the real-time values ​​and corresponding parameter ranges of the working parameters of each working component, infers and outputs the operation control curves of each working component under the current working condition to control the excavator to perform autonomous operation. When the operation time for each working condition reaches the set time, it determines whether to switch to the next working condition based on the pre-trained working condition adaptation switching function. The independent variables of the working condition adaptation switching function are the pressure change curve of the main pump pressure and the working time of each component under the current working condition. If the determination result is yes, the corresponding autonomous switching model is called to output the working condition switching control curves of each working component. After the working condition switching control curves are completed, the autonomous operation model for the next working condition is called to perform autonomous operation. If the determination result is no, the real-time values ​​of the working parameters of each working component are fine-tuned within the parameter range until the determination result of the working condition adaptation switching function is yes.

Images