A bucket wheel machine current-based material taking boundary identification system and communication method

Abstract

The application discloses a kind of based on bucket wheel machine current's material taking boundary identification system and communication method, belong to automation control technical field.The system includes current monitoring, pitch angle acquisition, dynamic baseline modeling, load disturbance analysis, flow prediction and boundary probability evaluation etc.Module.Based on real-time current and pitch angle, construct and dynamically update a no-load current baseline that changes with pitch angle;Fusion analysis the deviation characteristics of this baseline and real-time current and independently predicted material taking flow, the probability that bucket wheel machine is in material area, boundary area and no material area is evaluated through fusion logic, and control command is generated accordingly.The application overcomes the low precision defect of traditional fixed threshold method through dynamic baseline;Multi-source information fusion decision greatly improves the robustness and intelligent level of system, realizes accurate and reliable identification to material taking boundary under the premise of not relying on expensive external sensor.

Description

Technical Field

[0001] This invention relates to the field of automatic control, and specifically designs a material handling boundary identification system and communication method based on the current of a bucket wheel excavator. Background Technology

[0002] Bucket wheel stacker-reclaimers are core equipment in bulk material handling systems in ports, power plants, and mines, and their level of automation and operational efficiency directly impact the entire production chain. A crucial aspect of bucket wheel stacker operations is accurately determining the contact state between the bucket wheel and the material pile, i.e., real-time identification of the "material-containing zone," "boundary zone," and "material-free zone." This is essential for improving efficiency, reducing energy consumption, and minimizing equipment wear. Currently, there are several shortcomings in how this is achieved. The most traditional manual operation relies entirely on the operator's experience and visual judgment. This is not only labor-intensive but also prone to misjudgment due to fatigue and obstructed vision, especially at night or in harsh conditions such as dust, rain, or fog. This can lead to numerous instances of unproductive excavation or equipment collisions, making it difficult to ensure operational stability and optimization. To overcome the limitations of manual operation, the industry has developed automation solutions based on external sensors such as laser scanners, radar, or 3D vision. However, these solutions are not only expensive in terms of equipment and maintenance but also suffer from drastic performance degradation under common industrial conditions such as dust and moisture. Furthermore, they can only acquire the surface outline of the material pile and cannot detect the real-time collapse or other internal dynamics of the material. Another simpler approach to automation involves monitoring the current of the bucket wheel drive motor to determine boundaries. The basic logic is to use the current rise and fall caused by changes in the material handling load and set a fixed threshold for differentiation.

[0003] However, the fundamental flaw of this method lies in the instability of its identification benchmark. The no-load current of a bucket wheel excavator is not a constant value; it changes significantly with the pitch angle of the bucket wheel boom and is also affected by various factors such as equipment wear and power grid fluctuations. Therefore, using a single fixed threshold cannot adapt to complex dynamic working conditions, easily leading to identification errors, resulting in material residue or prolonged empty digging. Its identification accuracy and reliability are far from meeting the requirements of efficient automated operation. In summary, existing technologies have significant shortcomings in achieving accurate, reliable, and low-cost automated identification of the material handling boundary of a bucket wheel excavator. There is an urgent need for an intelligent boundary identification method that can overcome environmental interference, does not rely on expensive external sensors, and can dynamically adapt to changes in the equipment's own state. Summary of the Invention

[0004] To address the aforementioned problems in the existing technology, the present invention aims to provide a material handling boundary identification system based on bucket wheel excavator current, comprising:

[0005] The current monitoring module is used to collect real-time current data from the bucket wheel and belt motor.

[0006] The pitch angle acquisition module is used to acquire the pitch angle data of the bucket wheel excavator in real time.

[0007] The dynamic baseline modeling module, whose input is connected to the current monitoring module and the pitch angle acquisition module, is used to construct and dynamically update an no-load current baseline model that characterizes the no-load operation state of the equipment based on the received real-time current data and pitch angle data, and output the no-load current baseline.

[0008] The load disturbance analysis and feature extraction module has its input terminals connected to the current monitoring module and the dynamic baseline modeling module, respectively. It is used to analyze the deviation between the received real-time current data and the no-load current baseline, and extract disturbance features that characterize the material load.

[0009] The flow prediction module, whose input is connected to the current monitoring module and the pitch angle acquisition module, is used to fuse the received real-time current data, pitch angle data and other relevant operating parameters, and use an adaptive model to predict and output the real-time material handling flow.

[0010] The boundary probability assessment module, whose input terminals are connected to the load disturbance analysis and feature extraction module and the flow prediction module respectively, is used to combine the received disturbance features and the real-time material handling flow rate, and use fusion logic to assess and output the boundary probability of the bucket wheel excavator being in different material handling areas.

[0011] The boundary determination and dynamic adjustment module has its input end connected to the boundary probability evaluation module. It is used to make a determination based on the received boundary probability and generate an adjustment command.

[0012] The communication module is used to realize data transmission between the modules and send the adjustment instructions generated by the boundary determination and dynamic adjustment module to the external control system.

[0013] Furthermore, the boundary identification method includes step S1, in which the current monitoring module and the pitch angle acquisition module acquire and preprocess data to acquire the current of the bucket wheel motor, the current of the belt motor, and the pitch angle of the bucket wheel in real time; and the dynamic baseline modeling module generates a dynamic current reference that can characterize the no-load operation state of the current equipment at the corresponding pitch angle based on the real-time acquired current and pitch angle.

[0014] Step S2: The load disturbance analysis and feature extraction module calculates the difference between the real-time current and the dynamic current reference to extract disturbance features characterizing changes in material load. In parallel, the flow prediction module estimates the current material handling flow rate based on the operating parameters of the real-time current and pitch angle. The boundary probability assessment module integrates the load disturbance features and the estimated material handling flow rate to comprehensively assess the probability that the bucket wheel excavator is currently in the material-rich zone, the boundary zone, and the material-free zone.

[0015] In step S3, the boundary determination and dynamic adjustment module determines whether the bucket wheel excavator has reached or is close to the material collection boundary based on the boundary zone and material-free zone probabilities evaluated in step S2 and according to the preset confidence rules. Once the determination is successful, a corresponding adjustment control command is generated and sent to the external control system through the communication module.

[0016] Furthermore, the step of generating the dynamic current reference in step S1 includes,

[0017] Step S101: Based on historically collected no-load operating current data of the equipment at different pitch angles, an initial baseline model that can reflect the relationship between pitch angle and no-load current is constructed offline.

[0018] Step S102: During real-time operation, a recursive state estimation method is adopted. The state estimation method iteratively corrects the internal state or parameters of the initial baseline model based on the deviation between the model prediction value and the real-time measurement value, thereby generating and outputting a dynamic current reference that can accurately adapt to the current equipment state and operating conditions.

[0019] Furthermore, the step of extracting disturbance features characterizing changes in material load in step S2 includes:

[0020] Step S201: Calculate the difference between the real-time current and the dynamic current reference generated in step S1 to obtain a residual signal that only reflects load changes.

[0021] Step S202: A signal multi-scale decomposition technique is used to decompose the residual signal into a series of characteristic components at different time or frequency scales in order to capture load disturbances of different natures.

[0022] Step S203: Extract energy, entropy or singular value statistical indicators from the feature components respectively, and combine these indicators into a multi-dimensional feature vector to quantitatively characterize the current material load state.

[0023] Furthermore, the estimated material flow rate in step S2 It is generated through a dynamically weighted adaptive prediction model, the output of which satisfies the following relationship:

[0024] ,in: It is a nonlinear mapping function obtained through offline training; and These are the increments of the current of the bucket wheel motor and the belt motor relative to the dynamic current reference, respectively. This is the current pitch angle.

[0025] It is a historical state factor, representing a short period of time in the past. The cumulative effect of internal load characteristics or flow changes; this factor is used to dynamically adjust the internal weights or biases of the nonlinear mapping function.

[0026] The process of comprehensively evaluating probabilities in step S2 is achieved by using the load disturbance characteristics and the estimated material handling flow as two independent information sources and employing a fusion logic. The final confidence of the state is enhanced based on the consistency of the judgments of the two information sources on the same state. An uncertainty index characterizing the overall reliability of the current boundary identification result is quantified based on the degree of conflict between the judgments of the two information sources. Based on the enhanced confidence, the probability evaluation values ​​of the material area, boundary area, and material-free area are normalized and generated.

[0027] A communication method for identifying the material handling boundary based on bucket wheel excavator current, which generates process status assessment data by processing real-time sensor data and issues control commands accordingly, is characterized by the following steps:

[0028] Step Y1 involves classifying the data streams within the system into at least three types: real-time sensor data containing equipment load and location information, process status assessment data characterizing the risk level of system operation, and control commands for intervening in equipment operation; and establishing the correlation between the process status assessment data and the real-time sensor data transmission strategy.

[0029] Step Y2: Real-time analysis of the risk level indicators and / or assessment result uncertainty indicators contained in the process status assessment data; based on the values ​​of the indicators, dynamically adjust the transmission priority of the above three types of data streams through a preset mapping rule; when the risk level increases or the assessment result uncertainty increases, automatically increase the priority of real-time sensing data and control commands.

[0030] Step Y3: Set a dynamic change threshold for the real-time sensing data. Reporting is triggered only when the change in the sensing data exceeds the threshold. The dynamic change threshold is adjusted based on the risk level index in the process status assessment data. When the risk level increases, the threshold is automatically lowered to increase the frequency and sensitivity of data reporting.

[0031] Step Y4: At the edge nodes near the sensor, perform preliminary feature extraction or aggregation operations on the raw, high-frequency real-time sensor data to generate a feature summary containing key change information. When the system is running at low risk, prioritize transmitting the feature summary with a smaller data volume, while in a high-risk state, switch to transmitting more complete raw or lightly processed sensor data.

[0032] Furthermore, the method for determining the dynamic change threshold in step Y3 includes:

[0033] Step Y301: Extract risk level indicators and assessment result uncertainty indicators from the process status assessment data, and monitor the quality indicators of the current communication link.

[0034] Step Y302: Normalize the three indicators respectively, and assign dynamic weights to them according to the current core operating objectives of the system.

[0035] Step Y303: The weighted state indicators are integrated into a comprehensive adjustment factor through a preset nonlinear function.

[0036] Step Y304: Calculate the final dynamic change threshold based on a preset baseline threshold and the adjustment factor. The relationship is as follows: when the comprehensive adjustment factor indicates an increase in system risk or a decrease in communication quality, the dynamic change threshold is reduced accordingly.

[0037] Furthermore, the edge feature aggregation and transmission in step Y4 employs an adaptive aggregation depth strategy, which includes:

[0038] Multiple aggregation levels are defined, ranging from transmitting only raw data to transmitting highly condensed statistical features, forming an aggregation level sequence. Based on the risk level indicators in the process status assessment data, the current aggregation level to be executed is dynamically selected. When the risk level is low, a deep aggregation level is selected, transmitting only a small amount of statistical data such as mean and variance. As the risk level increases, the aggregation depth is gradually reduced, and intermediate layer features, such as signal segment energy and frequency domain peaks, which are richer but have a larger data volume, are transmitted instead. When the risk level reaches its highest, the system switches to a no-aggregation mode, directly transmitting the raw sensor data to ensure the highest information fidelity.

[0039] Furthermore, the method includes a dynamic communication mode selection mechanism, the dynamic selection mechanism comprising:

[0040] The system monitors the status of various available communication links and assesses their current transmission latency, bandwidth, and energy consumption costs. It acquires risk level indicators from the process status assessment data. Through a decision logic, it weighs the urgency of the current task against the cost and performance of different communication links, dynamically selecting the optimal transmission path and communication mode for different data streams. The decision rule is as follows: for high-priority control commands and sensor data in high-risk states, the link with the lowest latency and highest reliability is prioritized, even if its energy consumption cost is higher; while for non-critical data in low-risk states, the link with the lowest energy consumption cost is prioritized.

[0041] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0042] This invention overcomes the identification difficulties caused by changes in operating conditions in the traditional fixed threshold method by constructing a baseline model of no-load current that dynamically changes with the pitch angle. This model can accurately reflect the true no-load state of the equipment under different postures, thus more sensitively and accurately extracting the true material load information from the current signal, greatly improving the accuracy and reliability of boundary identification, and effectively avoiding no-load excavation and overload.

[0043] This invention significantly improves the robustness and intelligence of the system by fusing multi-source information for decision-making. It integrates current disturbance characteristics with independent material handling flow prediction results for evaluation, enhancing the confidence level of judgment through information consistency, and ultimately outputting the probability of each material handling area. Compared to single-source methods, this invention provides more comprehensive and robust decision-making, effectively resisting data noise interference, and achieving deep intelligent perception of the material handling status without adding costly external sensors. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of the system module structure of the present invention.

[0045] Figure 2 This is an exemplary flowchart of the boundary recognition method of the present invention.

[0046] Figure 3 This is an exemplary flowchart of the steps involved in generating a dynamic current reference according to the present invention.

[0047] Figure 4 This is an exemplary flowchart of a step in the present invention for extracting disturbance features.

[0048] Figure 5 This is an exemplary flowchart of the communication method of the present invention.

[0049] Figure 6 This is an exemplary flowchart of the steps for determining the dynamically changing threshold in this invention. Detailed Implementation

[0050] The present invention will be further described below with reference to specific embodiments.

[0051] like Figure 1 The diagram shown is a schematic representation of the module structure of a material handling boundary identification system based on bucket wheel excavator current provided in this embodiment, including:

[0052] The system comprises the following modules: a current monitoring module for real-time acquisition of current data from the bucket wheel and belt motor; a pitch angle acquisition module for real-time acquisition of pitch angle data from the bucket wheel excavator; a dynamic baseline modeling module, whose inputs are connected to the current monitoring and pitch angle acquisition modules, for constructing and dynamically updating an unloaded current baseline model representing the equipment's unloaded operating state based on the received real-time current and pitch angle data, and outputting the unloaded current baseline; a load disturbance analysis and feature extraction module, whose inputs are connected to the current monitoring and dynamic baseline modeling modules respectively, for analyzing the deviation between the received real-time current data and the unloaded current baseline, and extracting disturbance features representing the material load; and a flow prediction module, whose inputs are connected to the current monitoring and pitch angle acquisition modules. The system includes: an angle acquisition module, which integrates received real-time current data, pitch angle data, and other relevant operating parameters; an adaptive model, which predicts and outputs the real-time material handling flow rate; a boundary probability assessment module, whose inputs are connected to the load disturbance analysis and feature extraction module and the flow prediction module, which combines received disturbance features and real-time material handling flow rate to assess and output the boundary probability of the bucket wheel excavator in different material handling zones using fusion logic; a boundary determination and dynamic adjustment module, whose input is connected to the boundary probability assessment module, which determines the boundary probability based on the received boundary probability and generates adjustment commands; and a communication module, which enables data transmission between modules and sends the adjustment commands generated by the boundary determination and dynamic adjustment module to the external control system.

[0053] like Figure 2 The diagram shown is an exemplary flowchart of the boundary recognition method executed by the boundary recognition system in this embodiment, including:

[0054] Step S1 involves data acquisition and preprocessing by the current monitoring module and the pitch angle acquisition module, which acquires the current of the bucket wheel motor, the current of the belt motor, and the pitch angle of the bucket wheel in real time. Based on the real-time acquired current and pitch angle, the dynamic baseline modeling module generates a dynamic current reference that can characterize the no-load operating state of the current equipment at the corresponding pitch angle.

[0055] In step S2, the load disturbance analysis and feature extraction module calculates the difference between the real-time current and the dynamic current reference to extract disturbance features characterizing changes in material load. In parallel, the flow prediction module estimates the current material intake flow rate based on the operating parameters of the real-time current and pitch angle. The boundary probability assessment module integrates the load disturbance features and the estimated material intake flow rate to comprehensively assess the probability that the bucket wheel excavator is currently in the material zone, boundary zone, and no-material zone.

[0056] In step S3, the boundary determination and dynamic adjustment module determines whether the bucket wheel excavator has reached or is close to the material taking boundary based on the boundary zone and no-material zone probabilities evaluated in step S2 and according to the preset confidence rules. Once the determination is successful, the corresponding adjustment control command is generated and sent to the external control system through the communication module.

[0057] like Figure 3 The diagram shows the step of generating a dynamic current reference in step S1 of this embodiment, which includes:

[0058] Step S101: Based on historically collected no-load operating current data of the equipment at different pitch angles, an initial baseline model that can reflect the relationship between pitch angle and no-load current is constructed offline.

[0059] In step S102, during real-time operation, a recursive state estimation method is adopted. The state estimation method iteratively corrects the internal state or parameters of the initial baseline model based on the deviation between the model prediction value and the real-time measurement value, thereby generating and outputting a dynamic current reference that can accurately adapt to the current equipment state and operating conditions.

[0060] like Figure 4 The step S2 in this embodiment, which involves extracting disturbance features characterizing changes in material load, includes:

[0061] Step S201 calculates the difference between the real-time current and the dynamic current reference generated in step S1, obtaining a residual signal that reflects only load changes. In one embodiment, step S201 is the starting point for load disturbance feature extraction, aiming to separate the true fluctuations caused by material load changes from the original measurement signal. The system acquires the operating current data of the equipment in real time through current sensors connected to the equipment motor or other actuators, such as Hall current sensors or current transformers. This real-time current data is typically acquired at a high-frequency sampling rate, forming a time series. Simultaneously, the system acquires a dynamic current reference from step S1. This reference is typically a current prediction value obtained by modeling or learning from historical data under ideal, no-load, or stable load conditions. It reflects the inherent operating characteristics of the equipment and non-load-related current changes. The core processor or embedded controller calculates the difference between the real-time current and the dynamic current reference in real time, thereby obtaining a residual signal that reflects only load changes. This residual signal is a time series, eliminating current fluctuations caused by non-load factors such as equipment start-up, stopping, and acceleration, focusing on revealing deviations caused by material load.

[0062] Step S202 employs a multi-scale signal decomposition technique to decompose the residual signal into a series of characteristic components at different time or frequency scales to capture load disturbances of different natures. In one embodiment, step S202 involves performing a deep analysis on the residual signal generated in S201 to distinguish different types of load disturbances. The system's data processing unit, such as signal processing software running on an industrial PC or dedicated DSP chip, employs a multi-scale signal decomposition technique. This technique can decompose a single residual signal into a series of characteristic components at different time or frequency scales. Exemplary decomposition techniques include: Discrete Wavelet Transform (DWT), which decomposes the signal into approximate and detail components at different scales, corresponding to the low-frequency and high-frequency characteristics of the signal, respectively; Empirical Mode Decomposition (EMD), which adaptively decomposes the signal into intrinsic mode functions (IMFs), reflecting the oscillation modes of the signal at different time scales; or Variational Mode Decomposition (VMD), which decomposes the signal into multiple mode components with compact spectra. Through this decomposition, the system can capture load disturbances of different natures. For example, sudden impact loads may be reflected in high-frequency components, while slow accumulation of materials or changes in rheological properties may be reflected in low-frequency components or specific modes.

[0063] Step S203 involves extracting energy, entropy, or singular value statistical indicators from the feature components, and combining these indicators into a multidimensional feature vector to quantitatively characterize the current material load state. In one embodiment, step S203 quantifies and integrates the multi-scale feature components decomposed in S202. The system's data analysis module, running on the same processing unit, extracts statistical indicators such as energy, entropy, or singular values ​​from each decomposed feature component. The energy indicator reflects the strength or fluctuation amplitude of the component at a specific scale. The entropy indicator measures the complexity or randomness of the component. Singular value decomposition (SVD) is a method for analyzing matrix data and can be used to extract the intrinsic structural features of multidimensional data. These statistical indicators are then combined into a multidimensional feature vector to quantitatively characterize the current material load state. This multidimensional feature vector provides a compressed and highly informative representation that comprehensively reflects various aspects of material load changes, such as the presence of abnormal impacts, abnormal vibration patterns, or slow load accumulation. These quantitative features serve as direct inputs to subsequent fault diagnosis models or optimization algorithms.

[0064] The estimated material flow rate in step S2 It is generated through a dynamically weighted adaptive prediction model, the output of which satisfies the following relationship:

[0065] ,in: It is a nonlinear mapping function obtained through offline training; and These are the increments of the current of the bucket wheel motor and the belt motor relative to the dynamic current reference, respectively. This is the current pitch angle.

[0066] It is a historical state factor, representing a short period of time in the past. The cumulative effect of internal load characteristics or flow changes; this factor is used to dynamically adjust the internal weights or biases of the nonlinear mapping function.

[0067] The process of comprehensively evaluating probabilities in step S2 is achieved by using load disturbance characteristics and estimated material handling flow as two independent information sources and employing a fusion logic. The final confidence of the state is enhanced based on the consistency of the judgments of the two information sources on the same state. An uncertainty index characterizing the overall reliability of the current boundary identification result is quantified based on the degree of conflict between the judgments of the two information sources. Based on the enhanced confidence, probability evaluation values ​​for the material-containing area, boundary area, and material-free area are generated by normalization.

[0068] like Figure 5The above illustrates a communication method for identifying the material handling boundary based on the current of a bucket wheel excavator, provided in this embodiment. The method generates process status assessment data by processing real-time sensor data and issues control commands accordingly. The method includes the following steps:

[0069] Step Y1 involves classifying the data streams within the system into at least three types: real-time sensor data containing equipment load and location information, process status assessment data characterizing the risk level of system operation, and control commands used to intervene in equipment operation; and establishing the correlation between process status assessment data and real-time sensor data transmission strategies.

[0070] Step Y2: Real-time analysis of the risk level indicators and / or assessment result uncertainty indicators contained in the process status assessment data; based on the value of the indicators, dynamically adjust the transmission priority of the above three types of data streams through a preset mapping rule. When the risk level increases or the assessment result uncertainty increases, the priority of real-time sensing data and control commands is automatically increased.

[0071] Step Y3 sets a dynamic change threshold for real-time sensor data. Reporting is only triggered when the change in sensor data exceeds this threshold. The dynamic change threshold is adjusted based on the risk level index in the process status assessment data. When the risk level increases, the threshold is automatically lowered to increase the frequency and sensitivity of data reporting.

[0072] Step Y4: At the edge nodes near the sensor, perform preliminary feature extraction or aggregation operations on the raw, high-frequency real-time sensor data to generate a feature summary containing key change information. When the system is running at low risk, prioritize transmitting the feature summary with a smaller data volume, while in a high-risk state, switch to transmitting more complete raw or lightly processed sensor data.

[0073] like Figure 6 As shown, the method for determining the dynamic change threshold in step Y3 of this embodiment includes:

[0074] Step Y301: Extract risk level indicators and uncertainty indicators of assessment results from the process status assessment data, and monitor the quality indicators of the current communication link;

[0075] Step Y302: Normalize the three indicators respectively, and assign dynamic weights to them according to the current core operating objectives of the system.

[0076] Step Y303: The weighted state indicators are fused into a comprehensive adjustment factor through a preset nonlinear function.

[0077] Step Y304: Calculate the final dynamic change threshold based on a preset baseline threshold and adjustment factor. The relationship is as follows: when the comprehensive adjustment factor indicates an increase in system risk or a decrease in communication quality, the dynamic change threshold is reduced accordingly.

[0078] The edge feature aggregation and transmission in step Y4 employs an adaptive aggregation depth strategy, which includes:

[0079] Define multiple aggregation levels, ranging from transmitting only raw data to transmitting highly condensed statistical features, forming a sequence of aggregation levels;

[0080] Based on the risk level indicators in the process status assessment data, dynamically select the aggregation level to be executed at the current time;

[0081] When the risk level is low, select the deep aggregation level to transmit only a small amount of statistical data such as mean and variance.

[0082] As the risk level increases, the aggregation depth is gradually reduced, and instead, intermediate layer features such as signal segment energy and frequency domain main peaks, which are richer but have a larger data volume, are transmitted.

[0083] When the risk level reaches its highest level, the system switches to non-aggregation mode to directly transmit raw sensor data, ensuring the highest information fidelity.

[0084] The method includes a dynamic communication mode selection mechanism, which includes:

[0085] Monitor the status of various communication links currently available in the system and assess their current transmission latency, bandwidth, and energy consumption costs; obtain risk level indicators from process status assessment data;

[0086] Through a decision-making logic, the urgency of the current task is weighed against the cost and performance of different communication links, and the optimal transmission path and communication mode are dynamically selected for different data streams.

[0087] The decision-making rule is as follows: for high-priority control commands and sensor data in high-risk states, the link with the lowest latency and highest reliability is selected first, even if its energy consumption cost is higher; while for non-critical data in low-risk states, the link with the lowest energy consumption cost is selected first.

[0088] The above content is merely an example and illustration of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the scope defined by the invention, they should all fall within the protection scope of the present invention.

Claims

1. A material handling boundary recognition system based on bucket wheel excavator current, characterized in that, include: The current monitoring module is used to collect real-time current data of the bucket wheel and belt motor. The pitch angle acquisition module is used to acquire the pitch angle data of the bucket wheel excavator in real time. The dynamic baseline modeling module, whose input is connected to the current monitoring module and the pitch angle acquisition module, is used to construct and dynamically update an no-load current baseline model that characterizes the no-load operation state of the equipment based on the received real-time current data and pitch angle data, and output the no-load current baseline. The load disturbance analysis and feature extraction module has its input terminals connected to the current monitoring module and the dynamic baseline modeling module, respectively. It is used to analyze the deviation between the received real-time current data and the no-load current baseline, and extract disturbance features that characterize the material load. The flow prediction module, whose input is connected to the current monitoring module and the pitch angle acquisition module, is used to fuse the received real-time current data, pitch angle data and other relevant operating parameters, and use an adaptive model to predict and output the real-time material feeding flow. The boundary probability assessment module, whose input terminals are connected to the load disturbance analysis and feature extraction module and the flow prediction module respectively, is used to combine the received disturbance features and the real-time material handling flow rate, and use fusion logic to assess and output the boundary probability of the bucket wheel excavator being in different material handling areas. The boundary determination and dynamic adjustment module has its input end connected to the boundary probability evaluation module, and is used to make a determination based on the received boundary probability and generate an adjustment instruction. The communication module is used to realize data transmission between various modules and to send the adjustment instructions generated by the boundary determination and dynamic adjustment module to the external control system.

2. The material handling boundary identification system based on bucket wheel excavator current according to claim 1, wherein the boundary identification method is executed, characterized in that: include, Step S1: The current monitoring module and the pitch angle acquisition module acquire and preprocess data to acquire the current of the bucket wheel motor, the current of the belt motor, and the pitch angle of the bucket wheel in real time; the dynamic baseline modeling module generates a dynamic current reference that can characterize the no-load operation state of the current equipment at the corresponding pitch angle based on the real-time acquired current and pitch angle. Step S2: The load disturbance analysis and feature extraction module calculates the difference between the real-time current and the dynamic current reference to extract disturbance features characterizing changes in material load. In parallel, the flow prediction module estimates the current material handling flow rate based on the real-time current and pitch angle operating parameters; the boundary probability assessment module integrates the load disturbance characteristics and the estimated material handling flow rate to comprehensively assess the probability that the bucket wheel excavator is currently in the material zone, boundary zone, and no-material zone, respectively. Step S3: The boundary determination and dynamic adjustment module determines whether the bucket wheel excavator has reached or is close to the material taking boundary based on the boundary zone and no-material zone probabilities evaluated in step S2 and according to the preset confidence rules. Once the determination is successful, a corresponding adjustment control command is generated and sent to the external control system via the communication module.

3. The material handling boundary identification system based on bucket wheel excavator current according to claim 2, characterized in that: The step of generating the dynamic current reference in step S1 includes, Step S101: Based on historically collected no-load operating current data of the equipment at different pitch angles, construct an initial baseline model offline that can reflect the relationship between pitch angle and no-load current. Step S102: During real-time operation, a recursive state estimation method is adopted. The state estimation method iteratively corrects the internal state or parameters of the initial baseline model based on the deviation between the model prediction value and the real-time measurement value, thereby generating and outputting a dynamic current reference that can accurately adapt to the current equipment state and operating conditions.

4. The material handling boundary identification system based on bucket wheel excavator current according to claim 2, characterized in that: The step of extracting disturbance features characterizing material load changes in step S2 includes: Step S201: Calculate the difference between the real-time current and the dynamic current reference generated in step S1 to obtain a residual signal that only reflects load changes. Step S202: A signal multi-scale decomposition technique is used to decompose the residual signal into a series of characteristic components at different time or frequency scales in order to capture load disturbances of different natures. Step S203: Extract energy, entropy or singular value statistical indicators from the feature components respectively, and combine these indicators into a multi-dimensional feature vector to quantitatively characterize the current material load state.

5. The material handling boundary identification system based on bucket wheel excavator current according to claim 2, characterized in that: The estimated material flow rate in step S2 It is generated through a dynamically weighted adaptive prediction model, the output of which satisfies the following relationship: ,in: It is a nonlinear mapping function obtained through offline training; and These are the increments of the current of the bucket wheel motor and the belt motor relative to the dynamic current reference, respectively. This is the current pitch angle; It is a historical state factor, representing a short period of time in the past. The cumulative effect of internal load characteristics or flow changes; this factor is used to dynamically adjust the internal weights or biases of the nonlinear mapping function.

6. The material handling boundary identification system based on bucket wheel excavator current according to claim 2, characterized in that: The process of comprehensively evaluating probabilities in step S2 is achieved by treating the load disturbance characteristics and the estimated material handling flow as two independent information sources and employing a fusion logic. The final confidence of the state is enhanced based on the consistency of the judgments of the two information sources on the same state. An uncertainty index characterizing the overall reliability of the current boundary identification result is quantified based on the degree of conflict between the judgments of the two information sources. Based on the enhanced confidence, the probability evaluation values ​​of the material area, boundary area, and material-free area are normalized and generated.

7. A communication method for identifying the material handling boundary based on bucket wheel excavator current, which generates process status assessment data by processing real-time sensor data and issues control commands accordingly, characterized in that... The method includes the following steps: Step Y1: Classify the data streams within the system into at least three types: real-time sensor data containing equipment load and location information, process status assessment data characterizing the risk level of system operation, and control commands for intervening in equipment operation; and establish the correlation between the process status assessment data and the real-time sensor data transmission strategy. Step Y2: Real-time analysis of the risk level indicators and / or assessment result uncertainty indicators contained in the process status assessment data; Based on the values ​​of the indicators, dynamically adjust the transmission priority of the three types of data streams through a preset mapping rule; When the risk level increases or the assessment result uncertainty increases, automatically increase the priority of real-time sensing data and control commands. Step Y3: Set a dynamic change threshold for the real-time sensing data. Reporting is triggered only when the change in the sensing data exceeds the threshold. The dynamic change threshold is adjusted based on the risk level index in the process status assessment data. When the risk level increases, the threshold is automatically lowered to increase the frequency and sensitivity of data reporting. Step Y4: At the edge nodes near the sensor, perform preliminary feature extraction or aggregation operations on the raw, high-frequency real-time sensor data to generate a feature summary containing key change information. When the system is running at low risk, prioritize transmitting the feature summary with a small data volume, while in a high-risk state, switch to transmitting more complete raw or lightly processed sensor data.

8. The material handling boundary identification communication method based on bucket wheel excavator current according to claim 7, characterized in that: The method for determining the dynamic change threshold in step Y3 includes: Step Y301: Extract risk level indicators and assessment result uncertainty indicators from the process status assessment data, and monitor the quality indicators of the current communication link; Step Y302: Normalize the three indicators respectively, and assign dynamic weights to them according to the current core operating objectives of the system. Step Y303: The weighted state indicators are fused into a comprehensive adjustment factor through a preset nonlinear function. Step Y304: Calculate the final dynamic change threshold based on a preset baseline threshold and the adjustment factor. The relationship is as follows: when the comprehensive adjustment factor indicates an increase in system risk or a decrease in communication quality, the dynamic change threshold is reduced accordingly.

9. The material handling boundary identification communication method based on bucket wheel excavator current according to claim 7, characterized in that: The edge feature aggregation and transmission in step Y4 employs an adaptive aggregation depth strategy, which includes: Define multiple aggregation levels, ranging from transmitting only raw data to transmitting highly condensed statistical features, forming a sequence of aggregation levels; Based on the risk level indicators in the process status assessment data, dynamically select the aggregation level to be executed. When the risk level is low, select the deep aggregation level and only transmit statistical data, which includes the mean and variance. As the risk level increases, the aggregation depth is gradually reduced, and intermediate layer features are transmitted instead. When the risk level reaches its highest level, the system switches to non-aggregation mode to directly transmit raw sensor data, ensuring the highest information fidelity.

10. The material handling boundary identification communication method based on bucket wheel excavator current according to claim 7, characterized in that: The communication method includes a dynamic communication mode selection mechanism, which includes: The system monitors the status of various available communication links and assesses their current transmission latency, bandwidth, and energy consumption costs; it also obtains risk level indicators from the process status assessment data. Through a decision-making logic, the urgency of the current task is weighed against the cost and performance of different communication links, and the optimal transmission path and communication mode are dynamically selected for different data streams. The decision-making rule is as follows: for high-priority control commands and sensor data in high-risk states, the link with the lowest latency and highest reliability is selected first, even if its energy consumption cost is higher; while for non-critical data in low-risk states, the link with the lowest energy consumption cost is selected first.

Images