Cooperative control method for multifunctional electric control system of loading machine

Abstract

The invention relates to the technical field of engineering machinery electric control, discloses a cooperative control method for a multifunctional electric control system of a loading machine, and aims to solve the problems of information island, asynchronous control, action jittering and low energy efficiency of an electric control system of a traditional loading machine. The method comprises the following steps: constructing a multi-source working condition information fusion sensing layer, collecting engine, hydraulic pressure, transmission and vehicle body state data in real time, and inputting the data into a central cooperative controller; establishing a collaborative decision model based on task driving, and dynamically generating hydraulic flow distribution, a gearbox gear, an engine rotating speed and a braking response threshold value; synchronously issuing a control instruction through a high-speed bus to ensure that the time sequence deviation of each subsystem is less than 0.5 millisecond; nonlinear coupling relation online decoupling compensation is implemented, action jitter is suppressed, and energy matching is optimized; and closed-loop performance evaluation and neural network adaptive optimization are completed. According to the technical scheme, the working efficiency, the operation smoothness, the energy utilization rate and the system safety can be remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of electrical control technology for construction machinery, specifically a collaborative control method for a multi-functional electrical control system of a loader. Background Technology

[0002] With the continuous improvement of the intelligence level of construction machinery, loaders, as core equipment in earthmoving operations, are developing their electronic control systems towards multi-functional integration and collaborative control. Modern loaders are typically equipped with multiple subsystems, including hydraulic, transmission, braking, working devices, and auxiliary operations. These subsystems need to work together efficiently under complex working conditions to achieve precise control and energy-saving operation. However, the electronic control architecture of traditional loaders often adopts a decentralized control strategy, with each functional module operating independently and serious information silos. This makes it difficult to dynamically coordinate control parameters according to the task, thus restricting further improvements in overall machine operating efficiency and operational comfort.

[0003] The collaborative control of the loader's multi-functional electronic control system aims to optimize power distribution, ensure smooth action transitions, and provide adaptive response to operating conditions by uniformly scheduling the operating status of each subsystem. The core of this technology lies in establishing a cross-system state perception, decision-making linkage, and execution feedback mechanism to address the real-time control requirements under complex actions such as loading, unloading, steering, and traveling.

[0004] In existing technologies, the electronic control systems of loaders generally rely on preset logic rules or single sensor feedback for local adjustments, lacking the ability to fuse and analyze multi-source operating condition information (such as load changes, road slope, engine speed, and control lever commands), leading to system response lag or control conflicts. Simultaneously, inconsistent communication protocols and asynchronous control cycles between different functional modules limit the accuracy of coordinated control, easily resulting in motion jitter or energy waste. Furthermore, under dynamic load changes or complex operating scenarios, existing systems struggle to achieve rapid decoupling and compensation of nonlinear coupling relationships, severely impacting operational stability and safety. Therefore, a method for achieving efficient, precise, and adaptive coordinated control of the loader's multi-functional electronic control system is urgently needed. Summary of the Invention

[0005] The purpose of this invention is to provide a collaborative control method for a multi-functional electronic control system of a loader, which can effectively solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A collaborative control method for a multi-functional electronic control system of a loader includes the following specific steps: Step 1: Construct a multi-source working condition information fusion perception layer for the loader. By deploying various sensors on the engine, transmission, hydraulic pump, working device, and chassis, the engine speed, hydraulic system pressure, transmission gear status, bucket attitude angle, vehicle acceleration, and road slope information are collected in real time. The multi-source working condition information is then uniformly converted into standardized digital signals and input into the central collaborative controller. Step 2: Establish a task-driven collaborative decision-making model. The central collaborative controller identifies the current task type based on the operating handle command and dynamically generates collaborative control target parameters for each subsystem by combining the multi-source working condition information. The collaborative control target parameters include the hydraulic system flow distribution ratio, the gearbox target gear, the engine target speed, and the braking system response threshold. Step 3: Execute cross-system synchronous control command issuance. The central coordinating controller, based on the preset unified time base, aligns the coordinating control target parameters according to the control cycle of each subsystem, and synchronously issues them to the hydraulic control unit, transmission control unit, braking control unit, and working device control unit through the high-speed controller local area network bus. Step 4: Implement online decoupling compensation for nonlinear coupling relationships. During the execution of control commands, each subsystem control unit monitors the deviation between its output response and the target parameters in real time and feeds the deviation information back to the central coordinating controller. The central coordinating controller dynamically adjusts the control parameters of other related subsystems based on a preset coupling relationship matrix to achieve jitter suppression and energy loss minimization. The coupling relationship matrix is ​​updated using the following formula: in, For the first The coupling coefficient matrix at time step 1. Output the deviation vector for each subsystem. For the observation matrix, This is the gain matrix for the recursive least squares method; Step 5: Complete closed-loop collaborative performance evaluation and adaptive optimization. After each job cycle, the central collaborative controller constructs a collaborative performance evaluation function based on the execution data and job efficiency indicators of each subsystem, and uses a lightweight neural network model trained with historical data to fine-tune the parameters of the collaborative decision model online to improve the collaborative accuracy and response speed of subsequent jobs.

[0007] Preferably, the multi-source operating condition information fusion sensing layer in step 1 includes an engine speed sensor, a hydraulic pressure sensor, a tilt sensor, a six-axis inertial measurement unit, and a slope radar. The engine speed sensor has a sampling frequency of 100 Hz, the hydraulic pressure sensor has a range of 0 to 35 MPa and an accuracy of ±0.1 MPa, the six-axis inertial measurement unit has an output frequency of 200 Hz, and the slope radar has a detection range of ±30 degrees and a resolution of 0.1 degrees.

[0008] Preferably, the task types in step 2 include loading operations, unloading operations, combined turning and walking actions, and fixed-point stacking operations. The central collaborative controller has a built-in task recognition rule library, which matches the current task type by the displacement speed, direction combination, and duration characteristics of the operating handle, with a recognition accuracy of greater than or equal to 98%.

[0009] Preferably, the generation of collaborative control target parameters in step 2 adopts a hierarchical optimization strategy. The upper layer sets a global energy efficiency target based on the type of work task, and the lower layer dynamically adjusts the local control parameters of each subsystem according to the real-time load status. The hydraulic system flow distribution ratio is calculated in real time based on the bucket load torque, with a distribution accuracy of ±2%. The engine target speed is determined jointly based on the gearbox gear and hydraulic load, with an adjustment step of 50 revolutions per minute.

[0010] Preferably, in step 3, the high-speed controller local area network bus communication baud rate is 1 megabit per second, the control cycle of each subsystem control unit is uniformly set to 10 milliseconds, and the central coordinating controller synchronously triggers the issuance of instructions at the beginning of each control cycle to ensure that the timing deviation of each subsystem action is less than 0.5 milliseconds.

[0011] Preferably, in step 4, the coupling relationship matrix is ​​constructed by combining offline calibration and online identification. In the offline calibration stage, interaction data of each subsystem is collected under typical working conditions to establish an initial coupling relationship model. In the online identification stage, the coupling coefficient is updated using the recursive least squares method based on real-time deviation data, with an update period of 100 milliseconds.

[0012] Preferably, in step 4, motion jitter suppression is achieved by introducing a feedforward compensation term in the hydraulic control unit. The feedforward compensation term is dynamically calculated based on the change rate of the output torque of the transmission system, and the compensation response delay is less than 5 milliseconds. Energy loss is minimized by optimizing the power matching curves of the engine and the hydraulic pump, and the matching error is controlled within ±3%.

[0013] Preferably, the collaborative performance evaluation function in step 5 includes four dimensions: operation cycle time, fuel consumption, smoothness index, and system response overshoot. The weights of each dimension are dynamically adjusted according to the type of operation task. The lightweight neural network model adopts a three-layer fully connected structure with 16 nodes in the input layer, 32 nodes in the hidden layer, and 8 nodes in the output layer. The model inference time is less than 2 milliseconds.

[0014] Preferably, it also includes: setting a safety redundancy mechanism in the central collaborative controller, which automatically switches to a degraded collaborative mode based on historical data prediction when any subsystem communication is interrupted or the sensor fails, maintaining basic operation functions, and the operation efficiency in the degraded mode is not less than 85% of the normal mode.

[0015] Preferably, it also includes: establishing a collaborative control database containing more than or equal to 50,000 sets of operation scenarios. The database covers control parameters and performance indicators under different soil types, slope conditions, load levels and operating habits, and is used for offline training of neural network models and online performance evaluation benchmark generation.

[0016] Compared with the prior art, the beneficial effects achieved by the present invention are: 1. This invention achieves comprehensive and real-time perception of the loader's operating status by constructing a multi-source working condition information fusion perception layer, solving the problem of information silos in traditional systems; by establishing a task-driven collaborative decision-making model, it can dynamically generate collaborative control target parameters for each subsystem according to specific operational needs, significantly improving the pertinence and adaptability of the overall machine control; by executing cross-system synchronous control command issuance, it unifies the control timing of each subsystem, effectively avoiding action conflicts and energy waste caused by asynchronous control cycles; by implementing online decoupling compensation for nonlinear coupling relationships, it achieves active suppression of mutual interference between systems under complex working conditions, greatly improving the smoothness and stability of operational actions; 2. By completing closed-loop collaborative performance evaluation and adaptive optimization, the system possesses continuous learning and self-improvement capabilities, ensuring that collaborative control performance continuously improves over time. In summary, this invention significantly improves the loader's operating efficiency, operational comfort, energy efficiency, and system safety without increasing hardware costs, providing a system-level solution for intelligent control of construction machinery. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall technical solution architecture of the present invention; Figure 2 This is a schematic diagram of the core principle framework of the task-driven collaborative decision-making model in this invention; Figure 3This is a data flow and perception fusion logic framework diagram between the multi-source working condition information fusion perception layer and the central collaborative controller in this invention; Figure 4 This is a schematic diagram of the collaborative execution framework for online decoupling and compensation of cross-system synchronous control command issuance and nonlinear coupling relationship in this invention; Figure 5 This is a logical flowchart of the closed-loop collaborative performance evaluation and adaptive optimization mechanism in this invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1 To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0020] Currently, as loaders are core equipment in earthmoving operations, their electronic control systems are developing towards multi-functional integration and collaborative control. Modern loaders are typically equipped with multiple subsystems, including hydraulic, transmission, braking, working devices, and auxiliary operations. These subsystems need to work together efficiently under complex working conditions to achieve precise control and energy-saving operation. However, the traditional electronic control architecture of loaders often adopts a decentralized control strategy, with each functional module operating independently and exhibiting severe information silos. This makes it difficult to dynamically coordinate control parameters according to the work task, hindering further improvements in overall machine efficiency and operational comfort. To address these technical problems, this invention proposes a method for collaborative control of a loader's multi-functional electronic control system. This method involves constructing a multi-source working condition information fusion perception layer, establishing a task-driven collaborative decision-making model, issuing cross-system synchronous control commands, implementing online decoupling compensation for nonlinear coupling relationships, and completing closed-loop collaborative performance evaluation and adaptive optimization.

[0021] refer to Figure 1The overall technical architecture of the collaborative control method for a multi-functional electronic control system of a loader proposed in this invention includes a central collaborative controller, a multi-source working condition information fusion sensing layer, a hydraulic control unit, a transmission control unit, a braking control unit, a working device control unit, and a high-speed controller local area network bus. The central collaborative controller, as the decision-making core of the entire collaborative control system, is responsible for receiving standardized digital signals from the multi-source working condition information fusion sensing layer and performing core functions such as collaborative decision-making, synchronous command issuance, online decoupling compensation, and performance adaptive optimization. Each subsystem control unit interacts with the central collaborative controller in real time through the high-speed controller local area network bus to ensure the synchronization of control commands and the timeliness of feedback information.

[0022] In the aforementioned collaborative control method for a multi-functional electronic control system of a loader, step 1 involves constructing a multi-source working condition information fusion sensing layer for the loader. This layer collects real-time information on engine speed, hydraulic system pressure, transmission gear status, bucket attitude angle, vehicle acceleration, and road slope using various sensors deployed on the engine, transmission, hydraulic pump, working device, and chassis. This multi-source working condition information is then uniformly converted into standardized digital signals and input into the central collaborative controller. Specifically, the multi-source working condition information fusion sensing layer includes an engine speed sensor, a hydraulic pressure sensor, a tilt sensor, a six-axis inertial measurement unit, and a slope radar. An engine speed sensor, mounted on the engine flywheel housing, operates at a sampling frequency of 100 Hz and is used to monitor real-time crankshaft speed changes. Hydraulic pressure sensors, located at the main hydraulic pump outlet and the inlets of each actuator cylinder, have a range of 0 to 35 MPa and an accuracy of ±0.1 MPa, and are used to acquire the load status of the hydraulic system. A tilt sensor, mounted on the rear wall of the bucket, measures the bucket's attitude angle relative to the horizontal plane. A six-axis inertial measurement unit, fixed near the vehicle's center of gravity, outputs at a frequency of 200 Hz and is used to collect the vehicle's linear acceleration and angular velocity information in three-dimensional space. A slope radar, mounted on the lower front of the vehicle body, has a detection range of ±30 degrees and a resolution of 0.1 degrees, and is used to perceive the longitudinal and lateral slopes of the working surface in real time. All analog or digital signals output from the sensors are filtered, amplified, and isolated by a signal conditioning circuit, and then uniformly converted into standardized 16-bit digital signals via an analog-to-digital converter module. These signals are then transmitted to the central coordinating controller via the local controller's local area network bus. The central collaborative controller has an internal data preprocessing module that performs timestamp alignment, outlier removal, and data interpolation on the input signals to ensure the consistency and integrity of multi-source information over time. The data preprocessing module employs a sliding window mid-range filtering algorithm with a window length of 5 sampling points to suppress transient noise interference from the sensors. For data loss caused by brief communication interruptions, linear interpolation based on historical data is used for compensation, with interpolation errors controlled within the sensor's accuracy range.

[0023] In the aforementioned collaborative control method for a multi-functional electronic control system of a loader, step 2 involves establishing a task-driven collaborative decision-making model. The central collaborative controller identifies the current task type based on the operating handle commands and dynamically generates collaborative control target parameters for each subsystem by combining the multi-source working condition information. These collaborative control target parameters include the hydraulic system flow distribution ratio, the gearbox target gear, the engine target speed, and the braking system response threshold. Specifically, the central collaborative controller has a built-in task identification rule library that stores feature templates for four typical tasks: loading, unloading, steering and walking combined actions, and fixed-point stacking. The task identification process involves real-time acquisition of the displacement speed, direction combination, and duration characteristics of the operating handle, and dynamic time warping matching with the templates in the rule library, achieving an accuracy rate of ≥98%. For example, in loading operation identification, if the system detects that the operating handle is continuously moving upward in the vertical direction with a displacement speed greater than 0.3 meters per second, while there is no significant horizontal displacement, and the duration exceeds 2 seconds, it is determined to be a loading operation. The generation of collaborative control target parameters employs a hierarchical optimization strategy. The upper layer sets global energy efficiency targets based on the identified task types; for example, loading operations focus on improving digging force and cycle speed, while unloading operations emphasize smoothness and positioning accuracy. The lower layer dynamically adjusts the local control parameters of each subsystem based on real-time load conditions. The hydraulic system flow distribution ratio is calculated in real-time based on the bucket load torque, which is derived from the hydraulic pressure sensor readings and bucket geometric parameters, with a distribution accuracy of ±2%. The target engine speed is determined jointly based on the gearbox gear and hydraulic load, with an adjustment step of 50 revolutions per minute to ensure that power requirements are met while avoiding engine idling or overloading. (Reference) Figure 2 The core principle framework of the collaborative decision-making model includes a task identification module, a global goal setting module, a local parameter optimization module, and a parameter output interface. The modules interact at high speed through an internal data bus, and the decision cycle is 20 milliseconds.

[0024] In the aforementioned collaborative control method for a multi-functional electronic control system of a loader, step 3 involves issuing cross-system synchronous control commands. The central collaborative controller, based on a preset unified time reference, aligns the collaborative control target parameters according to the control cycles of each subsystem and synchronously issues them to the hydraulic control unit, transmission control unit, braking control unit, and working device control unit via a high-speed controller local area network (LAN) bus. Specifically, the LAN bus communication baud rate is 1 megabit per second, and the control cycle of each subsystem control unit is uniformly set to 10 milliseconds. The central collaborative controller integrates a high-precision real-time clock module driven by a temperature-compensated crystal oscillator, with a frequency stability better than ±10 ppm. At the beginning of each 10-millisecond control cycle (i.e., the time reference point), the central collaborative controller synchronously triggers the command issuance operation, encapsulating the collaborative control target parameters for each subsystem into standard LAN data frames and broadcasting them via the high-speed LAN bus. Upon receiving the data frames, each subsystem control unit immediately parses and executes the corresponding control commands, ensuring that the timing deviation of each subsystem's actions is less than 0.5 milliseconds. To ensure communication reliability, the data frames employ a cyclic redundancy check (CRC-16-CCITT) mechanism with a bit error rate (BER) of less than 10 to the power of -9. (Reference) Figure 4 The cross-system synchronous control command issuance framework includes the command scheduler of the central coordinating controller, the physical layer of the high-speed controller local area network bus, and the command receivers and executors of each subsystem control unit. The three work together to achieve microsecond-level command synchronization accuracy.

[0025] In the aforementioned collaborative control method for a multi-functional electronic control system of a loader, step 4 involves implementing online decoupling compensation for nonlinear coupling relationships. During the execution of control commands, each subsystem control unit monitors the deviation between its output response and the target parameters in real time and feeds this deviation information back to the central collaborative controller. The central collaborative controller dynamically adjusts the control parameters of other related subsystems based on a preset coupling relationship matrix, thereby suppressing motion jitter and minimizing energy loss. Specifically, the coupling relationship matrix is ​​constructed through a combination of offline calibration and online identification. In the offline calibration stage, interaction data from each subsystem is collected under typical working conditions (such as loading on flat ground or unloading on a 30-degree slope), and an initial coupling relationship model is established using system identification technology. In the online identification stage, the coupling coefficients are updated using a recursive least squares method based on real-time deviation data, with an update period of 100 milliseconds. The coupling relationship matrix is ​​updated using the following formula: in, For the first The coupling coefficient matrix at time step 1. Output the deviation vector for each subsystem. For the observation matrix, This is the recursive least squares gain matrix. Judder suppression is achieved by introducing a feedforward compensation term into the hydraulic control unit. This term is dynamically calculated based on the rate of change of the transmission system's output torque, which is jointly derived from the transmission output shaft torque sensor and engine speed. The compensation response delay is less than 5 milliseconds. Energy loss minimization is achieved by optimizing the power matching curves of the engine and hydraulic pump, with the matching error controlled within ±3%. For example, when a sudden increase in hydraulic system pressure is detected while the engine speed decreases, the system determines that this is a power mismatch caused by a sudden load change. It immediately adjusts the target engine speed and hydraulic pump displacement through the coupling relationship matrix to rematch their power outputs, thereby suppressing engine speed fluctuations and hydraulic system pressure shocks.

[0026] In the aforementioned collaborative control method for a multi-functional electronic control system of a loader, step 5 involves completing closed-loop collaborative performance evaluation and adaptive optimization. After each work cycle, the central collaborative controller constructs a collaborative performance evaluation function based on the execution data and work efficiency indicators of each subsystem. It then uses a lightweight neural network model trained with historical data to fine-tune the parameters of the collaborative decision-making model online, thereby improving the collaborative accuracy and response speed of subsequent operations. Specifically, the collaborative performance evaluation function includes four dimensions: work cycle time, fuel consumption, smoothness index, and system response overshoot. The weights of each dimension are dynamically adjusted according to the type of work task. For example, in shoveling operations, the weight of work cycle time is set to 0.4, the weight of fuel consumption is set to 0.3, the weight of smoothness index is set to 0.2, and the weight of system response overshoot is set to 0.1; while in fixed-point stacking operations, the weight of smoothness index is increased to 0.5. The lightweight neural network model employs a three-layer fully connected structure. The input layer has 16 nodes, receiving 16 dimensions of features including task type, soil hardness, slope, and initial load. The hidden layer has 32 nodes, using the ReLU activation function. The output layer has 8 nodes, corresponding to the adjustment of the 8 key parameters in the collaborative decision-making model. Model inference time is less than 2 milliseconds, ensuring real-time online fine-tuning. (Reference) Figure 5 The logical flow of the closed-loop collaborative performance evaluation and adaptive optimization mechanism includes four stages: data acquisition, performance scoring, model inference, and parameter update, forming a complete feedback loop.

[0027] Furthermore, this invention includes a safety redundancy mechanism in the central collaborative controller. When any subsystem communication is interrupted or a sensor fails, it automatically switches to a degraded collaborative mode based on historical data prediction to maintain basic operational functions. The operational efficiency in degraded mode is no less than 85% of the normal mode. The safety redundancy mechanism is implemented through a dual-core lockstep processor. The primary core is responsible for normal collaborative control logic, while the secondary core monitors the primary core's operating status and the health of the communication link in real time. Once an anomaly is detected, the secondary core immediately takes over control and invokes the degraded mode control strategy. The degraded mode control strategy uses historical data from the most recent 100 work cycles and employs a moving average prediction method to estimate missing sensor data or control parameters, ensuring the system can still operate safely in degraded mode.

[0028] Furthermore, this invention also includes establishing a collaborative control database containing at least 50,000 sets of operational scenarios. The database covers control parameters and performance indicators under different soil types, slope conditions, load levels, and operating habits, and is used for offline training of neural network models and online performance evaluation benchmark generation. The collaborative control database is deployed in the non-volatile memory of the central collaborative controller, and organizes data using a key-value pair structure. The key is the feature vector of the operational scenario, and the values ​​are the corresponding optimal control parameter set and performance indicators. The database supports incremental updates; after each operational cycle, if the collaborative performance evaluation function score is higher than the historical average, the data from this operational cycle is added to the database, continuously enriching the scenario coverage.

[0029] To illustrate the technical effects of this invention more specifically, a practical application scenario is constructed as follows: Loading hard clay in an open-pit mine. After the loader starts, the multi-source working condition information fusion sensing layer collects real-time data showing a road slope of 5 degrees (uphill), an initial bucket attitude angle of -15 degrees, and an engine speed of 1800 rpm. The control handle issues an upward lifting command, which the central coordinating controller recognizes as a loading operation. Based on the current working condition information, it generates the following coordinating control target parameters: the hydraulic system flow distribution ratio is adjusted to 70% for the lifting cylinder and 30% for the steering cylinder; the target gearbox gear is set to 1st gear; the target engine speed is increased to 2200 rpm; and the braking system response threshold is set to 0.3g. The command is synchronously sent to each subsystem via the high-speed controller local area network bus. During execution, the six-axis inertial measurement unit detects a longitudinal acceleration fluctuation of 0.2g in the entire vehicle, which is determined to be a coupling interference between the hydraulic system and the transmission system. The central coordinating controller immediately adjusts the hydraulic pump displacement and engine throttle opening according to the coupling relationship matrix to suppress vibration. After the operation cycle ended, the system recorded the cycle time as 18 seconds, fuel consumption as 0.8 liters, smoothness index as 0.92, system response overshoot as 5%, and collaborative performance evaluation score as 92 points. Based on this, the lightweight neural network model fine-tunes the parameters of the collaborative decision-making model to provide a better control strategy for the next similar operation.

[0030] Example 2 Based on Example 1, this invention can also be applied to the unloading of loose sand and soil at urban construction sites. In this scenario, the loader needs to complete precise unloading in a confined space, requiring extremely high smoothness of movement and positioning accuracy. The multi-source working condition information fusion perception layer detects that the road surface is horizontal concrete, the initial bucket attitude angle is 30 degrees, and the engine speed is 1500 rpm. The operating handle issues a slow downward tilting command, which the central collaborative controller recognizes as an unloading operation, generating collaborative control target parameters: the hydraulic system flow distribution ratio is adjusted to 90% for the tipping cylinder and 10% for the lifting cylinder, the gearbox target gear is set to neutral, the engine target speed is maintained at 1500 rpm to reduce noise, and the braking system response threshold is set to 0.1g to ensure vehicle stability. Due to the low hydraulic system flow requirement during unloading, the system automatically activates the engine idle speed optimization mode, fine-tuning the target speed to 1400 rpm to further reduce fuel consumption. During execution, the tilt sensor detected a bucket tilting angle deviation exceeding 0.5 degrees. The system then fine-tuned the opening of the flow control valve of the tipping cylinder through an online decoupling compensation mechanism to ensure the accuracy of the unloading position. After the operation cycle ended, the weight of the smoothness index in the collaborative performance evaluation function increased to 0.6, and the system scored 95 points. Based on this, the neural network model strengthened the fine flow control strategy under the unloading operation.

[0031] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

[0032] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for coordinating control of a plurality of electrical control systems of a loader, the method comprising: The method comprises the following specific steps: ​ Step 1: Constructing a loader multi-source working condition information fusion perception layer, collecting engine speed, hydraulic system pressure, gearbox gear state, bucket attitude angle, vehicle acceleration and road slope information in real time through multiple sensors arranged on the engine, gearbox, hydraulic pump, working device and vehicle chassis, and converting the multi-source working condition information into standardized digital signals to input a central cooperative controller; Step 2: Establishing a task-driven cooperative decision-making model, the central cooperative controller identifies the current task type according to the operation handle instruction, and dynamically generates cooperative control target parameters of each subsystem in combination with the multi-source working condition information, the cooperative control target parameters including hydraulic system flow distribution ratio, gearbox target gear, engine target speed and brake system response threshold; Step 3: Executing cross-system synchronous control instruction issuing, the central cooperative controller time-aligns the cooperative control target parameters according to a preset unified time reference according to the control period of each subsystem, and synchronously issues the cooperative control target parameters to the hydraulic control unit, transmission control unit, brake control unit and working device control unit through a high-speed controller local area network bus; Step 4: Implementing online decoupling compensation of nonlinear coupling relationship, each subsystem control unit monitors the deviation between the system output response and the target parameter in real time during the execution of the control instruction, and feeds back the deviation information to the central cooperative controller, the central cooperative controller dynamically adjusts the control parameters of other related subsystems based on a preset coupling relationship matrix, to realize action jitter suppression and energy loss minimization; Step 5: Completing closed-loop cooperative performance evaluation and adaptive optimization, the central cooperative controller constructs a cooperative performance evaluation function based on the execution data and work efficiency indicators of each subsystem after each work cycle, and uses a lightweight neural network model trained based on historical data to fine-tune the parameters of the cooperative decision-making model online to improve the cooperative accuracy and response speed of subsequent work.

2. The method of claim 1, wherein: The multi-source working condition information fusion perception layer comprises an engine speed sensor, a hydraulic pressure sensor, an inclination sensor, a six-axis inertial measurement unit and a slope radar; wherein the sampling frequency of the engine speed sensor is 100 Hz, the range of the hydraulic pressure sensor is 0-35 MPa, the accuracy is ±0.1 MPa, the output frequency of the six-axis inertial measurement unit is 200 Hz, and the detection range of the slope radar is ±30 degrees, the resolution is 0.1 degree.

3. The method of claim 1, wherein: The task types include loading operation, unloading operation, steering and walking combined action and fixed-point stacking operation; the central cooperative controller has a built-in task recognition rule library, which matches the current task type through the displacement speed, direction combination and duration characteristics of the operation handle, and the recognition accuracy is greater than or equal to 98%.

4. The method of claim 1, wherein: The generation of the cooperative control target parameter adopts a hierarchical optimization strategy, the upper layer sets a global energy efficiency target based on the type of the work task, and the lower layer dynamically adjusts local control parameters of each subsystem according to a real-time load state; wherein the hydraulic system flow distribution ratio is calculated in real time according to the load moment of the bucket, the distribution accuracy is plus or minus 2%, and the target engine speed is determined according to the gear position of the gearbox and the hydraulic load, and the adjustment step is 50 revolutions per minute.

5. The method of claim 1, wherein: The high-speed controller area network bus communication baud rate is 1 megabit per second, the control cycle of each subsystem control unit is uniformly set to 10 milliseconds, the central cooperative controller synchronously triggers the instruction issuing at the starting moment of each control cycle, and the motion time sequence deviation of each subsystem is less than 0.5 milliseconds.

6. The method of claim 1, wherein: The coupling relationship matrix is constructed by combining offline calibration and online identification; In the offline calibration stage, the interaction data of each subsystem is collected under typical working conditions to establish an initial coupling relationship model, and in the online identification stage, the coupling coefficients are updated by using the recursive least squares method based on real-time deviation data, and the update period is 100 milliseconds.

7. The method of claim 1, wherein: The motion jitter suppression is realized by introducing a feedforward compensation term in the hydraulic control unit, the feedforward compensation term is dynamically calculated according to the change rate of the transmission system output torque, and the compensation response delay is less than 5 milliseconds; the energy loss minimization is realized by optimizing the power matching curve of the engine and the hydraulic pump, and the matching error is controlled within plus or minus 3%.

8. The method of claim 1, wherein: The cooperative performance evaluation function includes four dimensions of work cycle time, fuel consumption, motion smoothness index and system response overshoot, and the weight of each dimension is dynamically adjusted according to the type of the work task; the lightweight neural network model adopts a three-layer full connection structure, the number of input layer nodes is 16, the number of hidden layer nodes is 32, and the number of output layer nodes is 8, and the model inference time is less than 2 milliseconds.

9. The method of claim 1, wherein: Further comprising setting a safety redundancy mechanism in the central cooperative controller, when any subsystem communication is interrupted or the sensor fails, automatically switching to a degraded cooperative mode based on historical data prediction, maintaining basic work functions, and the work efficiency in the degraded mode is not less than 85% of the normal mode.

10. The method of claim 1, wherein: Further comprising establishing a cooperative control database containing more than 50,000 work scenes, the database covers control parameters and performance indicators under different soil types, slope conditions, load levels and operation habits, and is used for offline training of the neural network model and online performance evaluation benchmark generation.

Images