AI algorithm-based unmanned intelligent operation method and system for bucket wheel machine of power plant

Abstract

This invention provides an AI-based method and system for unmanned intelligent operation of bucket wheel excavators in power plants, relating to the field of bucket wheel excavator technology. The method includes: acquiring power plant operating status information and bucket wheel excavator information; constructing a digital twin model of the power plant and bucket wheel excavator based on the power plant operating status information and bucket wheel excavator information; performing a correlation analysis between the power plant's fuel reserves and the bucket wheel excavator's transportation tasks based on the digital twin model to obtain an optimized transportation scheme; performing discrete event simulation processing based on the optimized transportation scheme to obtain efficiency simulation results; and performing optimization processing based on the efficiency simulation results, obtaining the operation scheme by setting an objective function and solving for the optimal solution. This invention, through modeling the future state prediction of power plant equipment, the state transition probability matrix of the bucket wheel excavator, and the collaborative working mode of the equipment, can monitor and dynamically adjust the operating status of the equipment in real time, thereby improving overall operational efficiency.

Description

Technical Field

[0001] This invention relates to the field of bucket wheel excavator technology, and more specifically, to a method and system for unmanned intelligent operation of a power plant bucket wheel excavator based on AI algorithms. Background Technology

[0002] In modern power plants, bucket wheel excavators serve as crucial material handling equipment, widely used for tasks such as fuel transport and ash handling. With increasing global energy demand and the expansion of power plant scale, the efficiency and reliability of bucket wheel excavators play a vital role in the stable operation of the entire power generation system. However, in traditional power plant bucket wheel excavator operation, equipment control relies primarily on mechanical systems and manual intervention. While this method meets basic needs in daily operation, its adaptability and efficiency face challenges as technological requirements increase. Current technologies often lack intelligent support for real-time equipment status monitoring, collaborative operation optimization, and energy consumption control. In particular, the collaborative operation between bucket wheel excavators and other equipment, and the planning of transport paths, largely rely on preset programs and manual adjustments, lacking real-time dynamic optimization and intelligent decision-making capabilities. Furthermore, existing systems typically monitor and maintain equipment based on a single physical model, failing to fully consider the complex interactions between equipment. This results in the system's inability to respond quickly to changes in equipment status or the environment, thus impacting overall production efficiency.

[0003] Based on the shortcomings of the existing technologies, there is an urgent need for an unmanned intelligent operation method and system for power plant bucket wheel excavators based on AI algorithms. Summary of the Invention

[0004] The purpose of this invention is to provide an AI-based method for unmanned intelligent operation of bucket wheel excavators in power plants, thereby addressing the aforementioned problems. To achieve this objective, the technical solution adopted by this invention is as follows:

[0005] Firstly, this application provides a method for unmanned intelligent operation of a power plant bucket wheel excavator based on an AI algorithm, including:

[0006] The system acquires power plant operating status information and bucket wheel excavator information. The power plant status information includes fuel reserves, boiler load, turbine status, and power generation capacity. The bucket wheel excavator information includes the bucket wheel excavator's location information, transport efficiency, material grabbing capacity, and equipment fault status.

[0007] Based on the power plant's operating status information and the bucket wheel excavator's information, a digital twin model of the power plant and bucket wheel excavator is constructed by performing trend analysis and status change modeling on the power plant's equipment status and the bucket wheel excavator's operating status.

[0008] Based on the digital twin model of the power plant-bucket wheel excavator, the correlation analysis between the power plant's fuel reserves and the bucket wheel excavator's transportation tasks is conducted. The transportation route is simulated through a path planning algorithm, and the loading and unloading efficiency and equipment coordination are evaluated in combination with real-time status data to obtain an optimized transportation plan.

[0009] Discrete event simulation processing is performed based on the optimized transportation scheme. By converting the transportation process into discrete events and combining real-time equipment status data for data format conversion and event-driven simulation, efficiency simulation results are obtained.

[0010] Based on the efficiency simulation results, optimization is performed by setting an objective function, which includes maximizing transportation efficiency, minimizing equipment failure rate, optimizing energy consumption and minimizing environmental impact. A multi-objective optimization algorithm is used to weight and combine the objectives, and the optimal solution is obtained to obtain the operation plan.

[0011] Secondly, this application also provides an unmanned intelligent operation system for power plant bucket wheel excavators based on AI algorithms, including:

[0012] The acquisition module is used to acquire power plant operating status information and bucket wheel excavator information. The power plant status information includes fuel reserves, boiler load, turbine status and power generation. The bucket wheel excavator information includes bucket wheel excavator location information, transportation efficiency, material grabbing volume and equipment fault status.

[0013] The modeling module is used to construct a digital twin model of the power plant and the bucket wheel excavator by performing trend analysis and state change modeling on the equipment status of the power plant and the operating status of the bucket wheel excavator based on the power plant's operating status information and the bucket wheel excavator information.

[0014] The analysis module performs a correlation analysis between the power plant's fuel reserves and the bucket wheel excavator's transportation tasks based on the power plant-bucket wheel excavator digital twin model. It simulates the transportation route through a path planning algorithm and evaluates the loading and unloading efficiency and equipment coordination by combining real-time status data, thereby obtaining an optimized transportation plan.

[0015] The simulation module is used to perform discrete event simulation processing based on the optimized transportation scheme. By converting the transportation process into discrete events and combining real-time equipment status data for data format conversion and event-driven simulation, efficiency simulation results are obtained.

[0016] The optimization module performs optimization processing based on the efficiency simulation results. By setting an objective function, which includes maximizing transportation efficiency, minimizing equipment failure rate, optimizing energy consumption and minimizing environmental impact, a multi-objective optimization algorithm is used to weight and combine the objectives, and the optimal solution is obtained to obtain the operation plan.

[0017] The beneficial effects of this invention are as follows:

[0018] This invention, through modeling the future state prediction of power plant equipment, the state transition probability matrix of bucket wheel excavators, and the collaborative working mode of equipment, enables real-time monitoring and dynamic adjustment of equipment operating status. This effectively optimizes equipment operation paths and modes, reduces unnecessary energy waste, and improves overall operational efficiency. By comprehensively considering the collaborative working mode of power plant equipment and bucket wheel excavators, and utilizing mechanical modeling and simulation technologies, this invention achieves efficient collaboration between equipment, ensuring maximum synergy among all devices and enhancing the adaptability and operational stability of bucket wheel excavators in complex operating environments. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of a method for unmanned intelligent operation of a power plant bucket wheel excavator based on an AI algorithm, as described in an embodiment of the present invention.

[0021] Figure 2 This is a schematic diagram of the structure of an unmanned intelligent operation system for a power plant bucket wheel excavator based on an AI algorithm, as described in an embodiment of the present invention.

[0022] Figure 3 This is a schematic diagram of the structure of an unmanned intelligent operation equipment for a power plant bucket wheel excavator based on an AI algorithm, as described in an embodiment of the present invention.

[0023] The diagram is labeled as follows: 800, an unmanned intelligent operation device for a power plant bucket wheel excavator based on AI algorithms; 801, processor; 802, memory; 803, multimedia component; 804, I / O interface; 805, communication component; 901, acquisition module; 902, modeling module; 903, analysis module; 904, simulation module; 905, optimization module. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0025] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0026] Example 1:

[0027] This embodiment provides a method for unmanned intelligent operation of bucket wheel excavators in power plants based on AI algorithms.

[0028] See Figure 1 The figure shows that the method includes steps S100 to S500.

[0029] Step S100: Obtain power plant operating status information and bucket wheel excavator information. Power plant status information includes fuel reserves, boiler load, turbine status and power generation. Bucket wheel excavator information includes bucket wheel excavator location information, transportation efficiency, material grabbing volume and equipment fault status.

[0030] Understandably, the core task of this step is to acquire real-time operational status information of the power plant and bucket wheel excavator. This data provides the foundation for subsequent status prediction, optimized scheduling, and intelligent decision-making. Power plant status information includes parameters such as fuel reserves, boiler load, turbine status, and power generation. This information helps assess the power plant's current operational capacity and resource utilization. For example, fuel reserves determine the power plant's continuous operating time, while boiler load and turbine status directly affect power generation efficiency and energy consumption. Power generation reflects the power plant's output capacity and production efficiency. This data not only helps predict the power plant's workload but also provides data support for equipment maintenance and optimization. Bucket wheel excavator information includes location information, transport efficiency, material grabbing volume, and equipment fault status. Location information determines the bucket wheel excavator's real-time position during transport; transport efficiency and material grabbing volume reflect the bucket wheel excavator's operational capacity and performance; and equipment fault status can provide early warnings of potential faults and downtime. Combining this information allows for accurate assessment of the collaborative operation between the bucket wheel excavator and power plant equipment, and timely adjustments to operational plans based on operational data to ensure efficient and stable system operation.

[0031] In practical applications, modern sensors, Internet of Things (IoT) technology, and real-time data acquisition systems are used to monitor and update this information in real time, ensuring the accuracy and timeliness of the data. Acquiring this information provides accurate input data for subsequent steps, enabling the system to react according to real-time conditions. By acquiring real-time status information of power plants and bucket wheel excavators, dynamic operation scheduling and resource optimization can be achieved, avoiding the lag of traditional manual intervention, improving the predictability and efficiency of equipment operation, and reducing production losses caused by equipment failure or energy waste.

[0032] Step S200: Based on the power plant's operating status information and the bucket wheel excavator's information, a digital twin model of the power plant and bucket wheel excavator is constructed by performing trend analysis and status change modeling on the power plant's equipment status and the bucket wheel excavator's operating status.

[0033] It should be noted that the digital twin model, through the combination of big data collection, real-time monitoring systems, and AI algorithms, can be updated in real time during equipment operation. It combines traditional equipment monitoring methods with advanced predictive analytics, enabling optimal choices for equipment operation and maintenance based on real-time data and future trends. This step, by constructing a digital twin model, achieves global intelligent optimization of the power plant and bucket wheel excavator system, accurately predicting equipment status, identifying potential faults in advance, thereby improving equipment operational stability, reducing downtime, and enhancing the overall system's resource utilization efficiency.

[0034] Step S300: Based on the digital twin model of power plant-bucket wheel excavator, conduct correlation analysis between power plant fuel reserves and bucket wheel excavator transportation tasks, simulate transportation routes through path planning algorithms, and evaluate loading and unloading efficiency and equipment coordination by combining real-time status data to obtain an optimized transportation plan;

[0035] First, correlation analysis models the relationship between power plant fuel reserves and bucket wheel excavator transport tasks. Information such as power plant fuel reserves and boiler load determines the demand for bucket wheel excavator transport tasks. For example, when fuel reserves decrease, the transport tasks of bucket wheel excavators need to be increased to ensure the power plant's continued operation; conversely, when fuel reserves are sufficient, the demand for transport tasks decreases. This correlation analysis uses multiple regression models, time series analysis, or machine learning methods to establish the link between power plant resource demand and bucket wheel excavator tasks, thereby providing accurate data input for route planning and transport scheduling.

[0036] Next, route planning involves simulating the transportation path using optimization algorithms (such as A* algorithm, Dijkstra's algorithm, or genetic algorithm) based on the power plant-bucket wheel excavator digital twin model and real-time status data. Route planning not only considers the shortest distance of the transportation route but also comprehensively takes into account factors such as equipment status, environmental conditions, and traffic flow in the work area. For example, in the event of equipment failure or maintenance, route planning needs to consider alternative routes or dynamically adjust the equipment's work allocation. Furthermore, real-time status data can dynamically update the route planning results to cope with unforeseen circumstances (such as equipment failure or weather changes). This route planning method ensures that the execution of transportation tasks can respond quickly to actual conditions, improving transportation efficiency.

[0037] Finally, the evaluation of loading and unloading efficiency and equipment coordination is conducted by combining real-time equipment data and transportation route information. This process involves real-time monitoring of equipment status (such as the grabbing capacity and transportation speed of the bucket wheel excavator) and, in conjunction with a collaborative work model, evaluating the collaborative work between the bucket wheel excavator and other equipment in the power plant (such as conveyor belts and unloading ports). Simulation techniques and performance analysis methods (such as Monte Carlo simulation or discrete event simulation) are used during the evaluation to simulate equipment coordination and loading and unloading efficiency under different paths, identify potential bottlenecks, and optimize them. Ultimately, by combining route planning and equipment coordination, an optimized transportation scheme is obtained to ensure the high efficiency and smoothness of the transportation process.

[0038] Step S400: Perform discrete event simulation processing based on the optimized transportation plan. By converting the transportation process into discrete events and combining real-time equipment status data for data format conversion and event-driven simulation, efficiency simulation results are obtained.

[0039] Discrete event simulation (DIS) transforms the entire transportation process into discrete events. By defining the time point and related conditions of each event, it simulates the actual operation of the transportation task. Each transportation link (such as loading, transporting, and unloading materials) is considered an independent event, which is discrete in time, with certain intervals between different events. During the simulation, each event is triggered according to preset rules or real-time input data, thus affecting the progress of the entire transportation process. For example, a component failure may cause a delay in the transportation task, or equipment failure may affect the operation path and transportation time. Through DIS simulation, the temporal relationships and state changes of each link in the transportation process can be simulated, thus providing data support for subsequent optimization schemes. The final efficiency simulation results are the evaluation results of different transportation schemes generated by the simulation system. These results show various efficiency indicators of the transportation process under different conditions, such as transportation time, equipment utilization, and energy consumption. Through simulation, potential bottlenecks can be identified, such as unreasonable transportation routes and delays caused by equipment failures, thus providing a basis for further adjustments to the optimization scheme. Simulation results can also help determine the performance of different optimization strategies in actual operation, providing a scientific basis for subsequent decision-making.

[0040] Step S500: Based on the efficiency simulation results, optimization is performed by setting an objective function, which includes maximizing transportation efficiency, minimizing equipment failure rate, optimizing energy consumption and minimizing environmental impact. A multi-objective optimization algorithm is used to weight and combine the objectives, and the optimal solution is obtained to get the operation plan.

[0041] It should be noted that the objective function is expressed as:

[0042] F total =w1·E transport -w2·F fault +w3·E energy -w4·E environment ;

[0043] Among them, F total Indicates the overall objective; E transport Indicates transportation efficiency; F fault Indicates equipment failure rate; E energy Indicates energy consumption; E environment The values ​​represent environmental impact; w1, w2, w3, and w4 represent the weights of each indicator.

[0044] Subsequently, the optimal solution obtained through a multi-objective optimization algorithm comprehensively considers the impact of all objectives, providing an optimal operational plan. This plan not only improves transportation efficiency and reduces equipment failures, but also effectively reduces energy consumption and environmental impact, ensuring the sustainability and efficiency of the overall operation.

[0045] Further, step S200 includes steps S210 to S240.

[0046] Step S210: Based on the power plant's operating status information, conduct equipment status trend analysis. By using the autoregressive integral moving average algorithm, analyze the trends of the power plant's fuel reserves, boiler load, turbine status, and power generation to obtain future predicted values ​​of the power plant's equipment status.

[0047] Specifically, firstly, trend analysis predicts future trends by modeling historical data on power plant operating status. Power plant operating status information includes fuel reserves, boiler load, turbine status, and power generation, all of which are crucial factors affecting equipment operating efficiency and energy management. Using the ARIMA algorithm, these time-series data can be modeled to capture the inherent patterns and time dependencies of the data, thereby predicting changes in equipment status over a future period. For example, by analyzing historical data on boiler load and fuel reserves, future energy demand and equipment load can be predicted, and resource allocation can be adjusted based on the prediction results.

[0048] The ARIMA model is a classic time series forecasting method suitable for data with significant time dependence and trend changes. It models the data using three key components: an autoregressive (AR) term, a differencing (I) term, and a moving average (MA) term. Specifically:

[0049] Autoregressive (AR): This assumes a linear relationship between the current value and the values ​​at previous time points.

[0050] Difference (I): By differentiating the time series, non-stationarity in the data is eliminated, making the mean and variance of the data stable.

[0051] Moving average (MA): Reduces the impact of noise on prediction results by smoothing out past data error terms.

[0052] By modeling power plant operating data (such as boiler load and fuel reserves) using the ARIMA model, predicted values ​​for each key parameter can be obtained over future time periods. These predictions not only reflect equipment operating trends but also provide a basis for subsequent dispatching decisions. For example, predicting future fuel demand and boiler load can effectively adjust power plant operating strategies, allowing for proactive resource allocation and equipment maintenance to ensure stable and efficient operation.

[0053] Step S220: Model the state change based on the bucket wheel excavator's operating state information. Model the state change of the bucket wheel excavator by constructing a hidden Markov model and simulate the operating state change of the bucket wheel excavator to obtain the state transition probability matrix of the bucket wheel excavator.

[0054] Understandably, a Hidden Markov Model (HMM) is a statistical model that assumes the system's state is not directly observable (hidden state), but can be inferred from certain observable measurable variables (observable state). In this embodiment, the bucket wheel excavator's operating state (such as normal operation, fault, standby, etc.) is considered the hidden state, while real-time information from monitoring the equipment (such as transport efficiency, grabbing volume, equipment fault status, etc.) is considered the observed state.

[0055] By constructing a Hidden Markov Model (HMM), the transition probabilities of a bucket wheel machine between different states can be described. Specifically, the HMM includes the following main components:

[0056] State space: Possible operating states of the bucket wheel excavator, such as "normal operation", "equipment failure", "standby", etc.

[0057] Transition probability matrix: Describes the probability of a bucket wheel excavator transitioning from one state to another. For example, the probability of a bucket wheel excavator transitioning from a "normal operation" state to a "equipment failure" state.

[0058] Observation probability: By monitoring the status of the equipment (such as the amount of food grasped, transportation efficiency, etc.), the probability of the observed result under each status is estimated.

[0059] The core task of Hidden Markov Models (HMMs) is to estimate the probabilities of state transitions using historical data and to predict future state changes based on these probabilities. By analyzing historical operating data of bucket wheel excavators (such as working status and related monitoring data), the transition probabilities are estimated using maximum likelihood estimation (MLE) or the Baum-Welch algorithm, resulting in a state transition probability matrix. This matrix represents the transition probabilities of the bucket wheel excavator between different states, providing data support for subsequent equipment management and optimized scheduling.

[0060] Furthermore, the process of generating the state transition probability matrix includes: modeling different operating states of the bucket wheel machine and extracting corresponding state transition records from historical data; using a learning algorithm of a hidden Markov model (such as a forward-backward algorithm) to calculate the transition probability between each state; and finally obtaining a transition probability matrix that provides a quantitative description of the transition between each state.

[0061] Ultimately, by constructing a Hidden Markov Model and obtaining the state transition probability matrix, the state change process of the bucket wheel excavator can be accurately simulated, providing predictive support for equipment maintenance, fault prediction, and scheduling. The state transition matrix not only helps assess the possible operating states of the bucket wheel excavator over a future period but also provides a basis for optimizing scheduling and resource allocation. For example, the probability of equipment failure can be predicted based on the transition probability, allowing for earlier maintenance and reduced downtime. By modeling state changes, the workload allocation of the equipment can be optimized, unnecessary equipment idleness can be reduced, and operational efficiency can be improved.

[0062] Step S230: Obtain historical data on the status of power plant equipment and the operating status of bucket wheel excavator, and conduct synergy effect analysis based on historical data. Use association rule mining to analyze the correlation between the status of power plant equipment and the status of bucket wheel excavator to obtain the equipment collaborative working mode.

[0063] Understandably, acquiring historical data involves recording the status of power plant equipment and bucket wheel excavators during long-term operation. Historical data for power plant equipment includes boiler load, fuel reserves, turbine status, and power generation, while historical data for bucket wheel excavators includes transport efficiency, material handling capacity, equipment fault records, and location information. This data is typically collected in real-time through sensors, data acquisition systems, and monitoring systems, forming time-series data. Comprehensive analysis of this historical data provides insights into equipment operating patterns, performance changes, and potential collaborative working mechanisms.

[0064] Next, synergy analysis explores how power plant equipment and bucket wheel excavators influence each other during operation by analyzing the interrelationships between them. Synergy typically manifests as the impact of changes in equipment status on the performance of another piece of equipment. For example, changes in the boiler load of a power plant directly affect the bucket wheel excavator's transport task scheduling, while a failure state of the bucket wheel excavator may affect the power plant's fuel consumption and power generation efficiency. By analyzing historical data, the interdependence and synergy patterns among these devices can be revealed.

[0065] Association rule mining analysis uses data mining techniques to reveal potential relationships between equipment states. Common association rule mining algorithms include the Apriori algorithm and the FP-growth algorithm, which can discover strong association rules between equipment states from large amounts of historical data. For example, the Apriori algorithm can identify the association rule that "when the boiler load is high, the transport efficiency of the bucket wheel excavator decreases." Through these association rules, the collaborative working patterns between the power plant equipment states and the bucket wheel excavator states can be derived, thus providing data support for the coordinated scheduling and optimization of equipment.

[0066] The collaborative working mode of equipment was derived by comprehensively analyzing the synergistic effects and the results of association rule mining, forming a model that reflects the collaborative work between equipment. This model describes how equipment influences each other and adjusts its tasks under different equipment states, ultimately achieving optimal resource and efficiency allocation.

[0067] Step S240: Based on the future predicted values ​​of the power plant equipment status, the state transition probability matrix of the bucket wheel excavator, and the equipment collaborative working mode, a convolutional neural network is used to extract features from the data of the power plant and the bucket wheel excavator in various dimensions, and an ensemble learning algorithm is used to fuse the models to construct a digital twin model of the power plant and the bucket wheel excavator.

[0068] In this step, the convolutional neural network (CNN) can automatically learn features from various dimensions of data related to power plant equipment (such as boiler load, fuel reserves, and power generation) and bucket wheel excavators (such as transport efficiency, equipment grabbing volume, and fault status). The CNN can capture the complex relationships between different dimensions, extract high-level features, and form efficient representations. Through multiple convolutional layers, pooling layers, and fully connected layers, the network can effectively process various data types (such as time series data, state information, and operating efficiency) and output a set of meaningful feature vectors. These features will serve as important inputs to the digital twin model for subsequent decision-making and optimization.

[0069] Next, ensemble learning algorithms are used to fuse the outputs of different models to improve the model's accuracy and robustness. Ensemble learning methods, such as XGBoost, LightGBM, or Random Forest, can reduce the bias and variance of a single model and improve the reliability of predictions by weightedly combining the predictions of multiple base models. Through ensemble learning algorithms, features extracted from convolutional neural networks and the outputs of other physical models (such as thermodynamic and mechanical models) are combined to form a comprehensive digital twin model of the power plant and bucket wheel excavator. This model can not only simulate and predict the behavior of the power plant and bucket wheel excavator under different operating conditions, but also optimize equipment collaborative working modes and resource scheduling strategies.

[0070] Further, step S220 includes steps S221 to S223.

[0071] Step S221: Perform state extraction processing based on the bucket wheel excavator's operating status information. Analyze the historical operating data of the bucket wheel excavator using a state switching point detection algorithm to obtain the switching point sequence of each state of the bucket wheel excavator.

[0072] Understandably, the purpose of state transition point detection is to identify moments when the state of equipment undergoes significant changes during operation. The operating states of a bucket wheel excavator include "normal operation," "equipment failure," and "standby," and transitions between these states are usually accompanied by sudden changes or trends in certain key indicators (such as material handling capacity, transport efficiency, and equipment temperature). Therefore, the goal of state transition point detection algorithms is to automatically identify these points of change, thereby dividing the complex equipment operating trajectory into clearer state intervals. Commonly used transition point detection algorithms include gradient-based detection algorithms, time-series analysis-based change point detection algorithms, and more complex statistical methods (such as the CUSUM algorithm and the Pelt algorithm). These algorithms can identify significant moments of state change based on the characteristics of the equipment state data and mark the transition points in the time-series data.

[0073] The generation of the switching point sequence is the output of the state switching point detection algorithm, which records the transition process of the bucket wheel excavator from one state to another. For example, the bucket wheel excavator may transition from a "normal operation" state to a "fault" state, or from a "standby" state back to "normal operation". These switching point sequences can provide the behavioral patterns of the bucket wheel excavator under different operating conditions, providing data support for subsequent modeling and analysis.

[0074] Step S222: Perform state transition modeling based on the switching point sequence. Model the state change process of the bucket wheel machine by constructing a hidden Markov model, and train the model using the maximum likelihood estimation algorithm to obtain the probabilistic framework of state transition.

[0075] It should be noted that the probabilistic framework for state transitions describes the transition relationships between different states and the probability of each transition. The probabilistic framework obtained through training a Hidden Markov Model can provide a reliable basis for subsequent decision support.

[0076] Step S223: Perform approximate probability calculation based on the probabilistic framework of state transition. By statistically analyzing the transition frequencies between states and normalizing them, the state transition probability matrix of the bucket wheel machine is obtained.

[0077] Understandably, the task of probabilistic probability calculation is to statistically analyze the frequency of transitions from one state to another. This process requires in-depth analysis of the bucket wheel excavator's historical operating data (e.g., material handling capacity, transport efficiency, equipment failure status, etc.). By analyzing the sequence of state transition points in historical data, common state transitions of the bucket wheel excavator under different operating modes can be identified. For example, if the bucket wheel excavator frequently transitions from "normal operation" to "equipment failure" within a certain period, then it can be inferred that the frequency of such transitions is relatively high. Based on this statistical analysis, the stability and reliability of the equipment under specific conditions can be understood.

[0078] Normalization converts the frequency of state transitions into probabilities. This process ensures that the sum of the probabilities of all state transitions is 1, thus forming a standardized model that can be used for further analysis and optimization. The normalized state transition probability matrix reflects the likelihood of transitions between different states. This information can help the system identify potential failure risks in advance and schedule spare parts and maintenance resources in a timely manner, thereby avoiding unexpected downtime and efficiency losses.

[0079] Further, step S240 includes steps S241 to S244.

[0080] Step S241: Based on the future predicted values ​​of the power plant equipment status, perform physical process modeling. By combining the thermodynamic model with the energy balance equation, simulate the energy flow and efficiency changes of the equipment in the future time period to obtain the predicted results of the thermodynamic dynamic changes of the power plant equipment.

[0081] It should be noted that physical process modeling involves mathematically modeling the thermodynamic processes of power plant equipment (such as boilers and turbines) to better predict their future operating states. The operation of power plant equipment involves complex heat energy conversion and flow. Key equipment, such as the heat input of boilers and the energy output of turbines, is affected by various factors, including fuel type, load demand, and equipment health. To accurately predict the future thermodynamic state of the equipment, thermodynamic models, such as energy-conservation-based thermodynamic analysis models, are needed to describe how heat energy flows within the equipment and is converted into useful work. These models are typically modeled using energy conservation equations (such as the first law), i.e., the relationship between the equipment's input energy (such as fuel) and output work. The formula for predicting the thermodynamic dynamic changes of power plant equipment is:

[0082]

[0083] Where t represents time; U(t) represents the energy stored in the device at time t; Q(t) represents the heat input at time t; and W(t) represents the work output at time t. h represents the mass flow rate of the fluid at time t; in (t), h out η(t) represents the enthalpy of the fluid flowing into and leaving the equipment, respectively; η(t) represents the thermal efficiency of the equipment at time t.

[0084] Step S242: Model the bucket wheel excavator based on its state transition probability matrix and equipment collaborative working mode. Through mechanical modeling and mechanical simulation, simulate the motion state and mechanical collaborative effect of the bucket wheel excavator to obtain the simulation results of the mechanical motion state of the bucket wheel excavator.

[0085] First, mechanical modeling involves modeling the motion of a bucket wheel excavator using mechanical principles (such as rigid body dynamics and dynamic equations). The operation of a bucket wheel excavator involves the interaction of multiple mechanical components, such as the drive system, conveyor belt, and grab bucket. The interactions between these components directly affect its working efficiency and performance. The goal of the mechanical model is to describe the motion state of these components, including the force and torque transmission between them. For example, the torque of the drive system is transmitted to the grab bucket, which then transports the material to a designated location. The mechanical relationships involved in this process can be modeled using rigid body mechanics and dynamic equations. Through these mechanical models, it is possible to understand and predict the behavior of the bucket wheel excavator under different operating conditions, especially its motion state during different operational stages such as material grabbing, transporting, and unloading.

[0086] Next, mechanical simulation uses computer simulation technology to simulate the motion process and synergistic effects of the bucket wheel excavator under different conditions. Mechanical simulation can transform a mechanical model into an operable computational model, simulating the actual motion of the bucket wheel excavator during operation. For example, using finite element method (FEM) or multibody dynamics (MBD) simulation tools, each component of the bucket wheel excavator is modeled, and simulations are performed based on real-time equipment status data (such as equipment failure status, transport efficiency, etc.) to obtain the motion trajectory and mechanical effects under different conditions. This process can simulate complex situations that may occur during equipment operation, such as load changes, equipment failures, speed changes, etc., and predict how these factors affect the overall operating efficiency of the bucket wheel excavator.

[0087] Equipment collaborative operation mode refers to the synergistic effect between the bucket wheel excavator and other equipment in the power plant (such as conveyor belts, transport vehicles, unloading equipment, etc.). In this process, changes in the bucket wheel excavator's status (such as transport efficiency, grabbing capacity, etc.) affect the load and efficiency of other equipment, while the operating status of other equipment also feeds back to the bucket wheel excavator's operating efficiency. By introducing a model of equipment collaborative operation into the simulation, the interaction between equipment can be taken into account, and the collaborative effects under different equipment operating states can be simulated, thereby optimizing task allocation and operation paths among equipment.

[0088] This step involves mechanical modeling and simulation by comprehensively considering the state transition probability matrix of the bucket wheel excavator and the equipment's collaborative working mode. The resulting mechanical motion simulation results accurately describe the bucket wheel excavator's motion behavior under different working conditions and reveal the efficiency changes of the equipment during collaborative work. These results provide data support for subsequent job scheduling, optimized path planning, and equipment maintenance decisions. Specifically, the motion equations involved in this embodiment are as follows:

[0089] The equation of motion for a bucket wheel excavator:

[0090]

[0091] Synergistic effect model equation:

[0092]

[0093] Coupling equations of the dynamic model:

[0094]

[0095] Where M represents the mass matrix, which is used to describe the inertia of each component in the system; Represents the acceleration vector; K(x) represents the damping matrix; K(x) represents the stiffness matrix; F ext This indicates the external load experienced by the bucket wheel excavator during operation; F int This represents the interaction forces between the components; n represents the number of components in the system; i represents the sequence number of the component currently participating in the cooperative action; x i x represents the state vector of the i-th component; j F represents the state vector of the j-th component; coupling,i M represents the cooperative force of the i-th component; eff This represents the effective quality matrix, i.e., the quality impact after considering synergistic effects; This represents the effective damping matrix, reflecting the additional drag under synergistic effects; K eff (x) represents the effective stiffness matrix, which takes into account the stiffness changes caused by collaborative work.

[0096] Step S243: Based on the thermodynamic dynamic change prediction results and the mechanical motion state simulation results, feature extraction is performed. Spatiotemporal information is extracted from the equipment status, energy efficiency and mechanical motion path dimensions of power plant equipment and bucket wheel excavator through a convolutional neural network to obtain feature vectors.

[0097] Understandably, thermodynamic dynamic change prediction results primarily describe the energy flow, efficiency changes, and future state of the equipment (e.g., changes in boiler load, fuel consumption, and power generation predictions), while mechanical motion state simulation results focus on the physical motion processes of the equipment, including the grabbing capacity, transport efficiency, and equipment state changes of the bucket wheel excavator. Combining these two types of information allows for a comprehensive description of the equipment's energy efficiency and physical motion behavior under different states, thus providing data support for subsequent feature extraction and optimization decisions.

[0098] Next, a Convolutional Neural Network (CNN) is used to extract spatiotemporal features from this multidimensional data. CNNs are highly capable in image processing and time-series data analysis, extracting high-level features from both spatial (e.g., different dimensions of equipment state) and temporal (e.g., changes in equipment state over time) aspects. In this step, the CNN automatically extracts effective spatiotemporal information from the state data of the power plant equipment and bucket wheel excavator through multiple convolutional and pooling layers. This spatiotemporal information helps capture the temporal dependencies and spatial features between equipment states, understand how the equipment changes in time and space, and form a comprehensive representation of equipment operation.

[0099] In convolutional neural network models, spatiotemporal information includes not only the current state of the equipment but also the influence of its historical states on the current state. For example, historical energy efficiency changes and the movement path changes of bucket wheel excavators can be extracted into features through convolution operations, forming a set of feature vectors. These feature vectors contain information across multiple dimensions, including equipment state, energy efficiency, and mechanical movement paths, providing detailed input data for subsequent digital twin modeling, optimization decisions, and fault prediction. The resulting feature vectors contain multi-dimensional, spatiotemporal information, providing foundational data for equipment state prediction and collaborative operation optimization. These feature vectors not only exhibit temporal correlation but also reflect the complex relationships between equipment states.

[0100] Step S244: Perform model fusion and optimization based on feature vectors. By using an ensemble learning algorithm, the output results of the thermodynamic model and the mechanical motion model are comprehensively optimized to obtain a digital twin model of the power plant-bucket turbine.

[0101] Ensemble learning is a technique that improves the overall performance of a model by combining the predictions of multiple models. Common ensemble learning methods include random forests, gradient boosting trees (GBDT), and XGBoost. In the digital twin model of a power plant-bucket wheel excavator, ensemble learning algorithms are used to fuse the outputs of a thermodynamic model and a mechanical motion model. These two models model the operating state of the equipment based on thermodynamic and mechanical principles, respectively. The thermodynamic model focuses on predicting energy flow and efficiency changes, while the mechanical motion model focuses on the physical motion state of the equipment and mechanical synergistic effects. By introducing ensemble learning algorithms, the advantages of these two models can be combined, reducing the error of a single model and improving the accuracy of predictions.

[0102] Model fusion involves taking the predictions from thermodynamic and mechanical motion models as inputs and weighting or training them using ensemble learning algorithms to allow the strengths of the two models to complement each other. For example, the XGBoost algorithm can train multiple decision trees and assign different weights to each, resulting in more accurate predictions and avoiding the limitations of over-reliance on any single model. Ensemble learning algorithms can handle the diversity and complexity of models, ensuring that the outputs of different models complement each other while maintaining accuracy, thus improving the robustness of the overall system.

[0103] Optimization is a core part of ensemble learning. In this step, the performance of the power plant-bucket wheel excavator digital twin model is optimized by weighted combination of results from different models. This process involves fine-tuning the model output to maximize prediction accuracy, reduce systematic errors, and improve the system's practical application effectiveness. For example, during optimization, the model weights may be dynamically adjusted based on the actual operating status of the equipment (such as equipment health status and energy consumption) to ensure that the digital twin model can adapt to changes in the power plant's operation in real time, making intelligent adjustments and predictions.

[0104] The resulting digital twin model of the power plant-bucket wheel excavator integrates information from both thermodynamics and mechanical motion, comprehensively predicting the equipment's operating status and efficiency changes. This model not only reflects the physical operating state of the equipment but also considers its energy efficiency and mechanical synergistic effects, providing strong data support for subsequent intelligent scheduling, fault prediction, and operation optimization. Through the digital twin model, the power plant can monitor equipment operating status in real time, predict potential problems in advance and make timely adjustments, reducing resource waste, improving equipment utilization, and ensuring the efficient and stable operation of the system.

[0105] Further, step S300 includes steps S310 to S340.

[0106] Step S310: Based on the equipment status information in the power plant-bucket wheel excavator digital twin model, perform transportation task demand prediction processing, and use the long short-term memory network algorithm to analyze the fuel reserve and boiler load change trends to obtain the expected transportation demand.

[0107] Specifically, the application of the Long Short-Term Memory (LSTM) network model is achieved through the following steps:

[0108] Step S311, Data Input: Input historical data (such as fuel reserves, boiler load, etc.) into the Long Short-Term Memory Network Model. This data is presented in time series form, and the Long Short-Term Memory Network Model can learn the time dependencies in it.

[0109] Step S312, Feature Learning and Memory: Long Short-Term Memory (LSTM) networks control information transmission through their unique gating mechanisms (input gate, forget gate, output gate), enabling them to automatically learn and remember long-term dependencies related to transportation tasks. For example, the relationship between boiler load changes and fuel reserve changes can be automatically captured and utilized by LTM networks during prediction.

[0110] Step S313, Demand Forecasting: Through training with a Long Short-Term Memory (LSTM) network, the model can predict transportation demand over a future period based on historical input data. For example, it can predict how much transportation will be needed to meet boiler load demand in the next few hours, days, or weeks.

[0111] Finally, the projected transportation demand is a prediction of future transportation tasks given by the Long Short-Term Memory Network model after analyzing the changing trends of fuel reserves and boiler load. This demand reflects the amount of transportation tasks that the bucket wheel excavator will need to complete in the future, directly affecting the allocation of transportation resources and the operation schedule of the bucket wheel excavator.

[0112] Step S320: Based on the transportation demand and the bucket wheel excavator's operating capacity, the task allocation is processed. The task quantity of each bucket wheel excavator is allocated through a mixed integer programming algorithm to obtain the specific operation plan of the bucket wheel excavator.

[0113] It should be noted that Mixed Integer Programming (MIP) is an optimization algorithm widely used in resource allocation and scheduling problems. MIP ensures the operability of task allocation by defining decision variables as integers (e.g., each bucket wheel excavator can only complete an integer number of tasks), and optimizes task allocation through constraints and an objective function. Specifically, in this step, the goal of MIP is to optimally allocate transportation tasks to meet transportation demand, while considering constraints such as the bucket wheel excavator's operating capacity and operating time. The objective function of MIP includes the following aspects:

[0114] Maximize transportation efficiency: Ensure that the task assignments of each bucket wheel excavator can be completed efficiently, reducing the time spent in idle or inefficient operation.

[0115] Minimize task delays: Ensure transportation tasks are completed within the stipulated time and reduce delays caused by work imbalances or improper resource scheduling.

[0116] Constraints include the operating capacity limits of the bucket wheel excavator (such as the maximum transport capacity of each bucket wheel excavator), operating time limits (such as restrictions on operating periods), and equipment status constraints (such as equipment failure or maintenance).

[0117] By combining transportation demand with the operating capacity of bucket wheel excavators, a mixed-integer programming algorithm can solve for the workload that each bucket wheel excavator should undertake. The task allocation result generates a specific work plan for each bucket wheel excavator, ensuring that the load of each piece of equipment matches its capacity and maximizing the overall efficiency of the system.

[0118] Step S330: Perform path planning based on the specific work plan. Simulate the transport route of the bucket wheel excavator using a dynamic path planning model, and obtain the optimal transport route by combining equipment status changes and environmental conditions.

[0119] Understandably, dynamic route planning models are based on shortest path algorithms in graph theory (such as A*, Dijkstra's algorithm, and genetic algorithms), combined with dynamically updated real-time data (such as equipment status and environmental conditions) to ensure high adaptability of route planning in changing environments. Specifically, these models dynamically evaluate the transportation efficiency of different routes while considering factors such as workload and time windows, and adjust routes based on the current operating status of equipment (such as faults or load). Route planning models need to handle the impact of equipment health status, workload, and external environmental factors (such as weather and traffic) on transportation routes in real time to ensure that each selected transportation route is optimal.

[0120] Step S340: Evaluate loading and unloading efficiency and process equipment collaborative work based on the optimal transportation route. By analyzing the efficiency of different routes and collaborative operation schemes, an optimized transportation scheme is obtained.

[0121] It should be noted that this method not only optimizes transportation routes but also reduces resource waste, lowers the risk of equipment failure, and maximizes equipment utilization through precise scheduling of equipment collaborative operations. Furthermore, the optimized transportation plan can dynamically adapt to changes in equipment status and environmental factors, providing more flexible and intelligent operation management and driving power plants towards greater efficiency and intelligence.

[0122] Further, step S400 includes steps S410 to S430.

[0123] Step S410: Perform event definition processing based on the optimized transportation plan. By converting transportation tasks, loading and unloading processes, and equipment operating status into discrete events and assigning them timestamps, a discrete event sequence is obtained.

[0124] Understandably, the purpose of event definition processing is to transform transportation tasks, loading and unloading processes, and equipment operating states into discrete events. In discrete event simulation, an event is a specific moment when the system state changes, usually corresponding to the completion of a specific operation or task. Therefore, each transportation task, loading and unloading process, and change in equipment operating state can be considered an event. For example, a bucket wheel excavator completing a material grabbing task, or a conveyor belt completing an unloading action, can be considered an "event" in the system. Changes in equipment operating state (such as changing from "normal operation" to "fault") can also be considered an event. In this step, these tasks and processes are modeled and transformed into discrete event objects. Next, timestamp assignment marks the timing of each event. A timestamp represents the specific time when an event occurs, and it is a crucial element in discrete event simulation. By assigning a timestamp to each event, it can be ensured that events occur in the actual time sequence and that the evolution of equipment states can be simulated. For example, a fault event of a piece of equipment may occur during transportation; the timestamp of this event will be marked as the specific time when the fault occurred. Through timestamps, the order of occurrence of each event can be precisely controlled, ensuring the time consistency of the simulation process. A discrete event sequence is an ordered set consisting of all the aforementioned events and their timestamps. This sequence reflects the temporal relationships of different events occurring in the system, such as the start and end of a transportation task and changes in equipment status. The discrete event sequence provides complete time-series data for subsequent simulation analysis. By analyzing these events, the system's performance under different operational schemes can be evaluated, such as the execution efficiency of transportation tasks and the operating status of equipment.

[0125] Step S420: Perform event triggering processing based on the digital twin model of the power plant-bucket wheel excavator. By combining the changes in equipment workload, operating environment and transportation route, dynamically adjust the timing of event occurrence to obtain a dynamic simulation scenario.

[0126] First, event-triggered processing determines in real-time whether certain critical events are triggered by monitoring multiple variables in the system (such as changes in equipment workload, operating environment, and transportation routes). For example, when the equipment workload exceeds a preset threshold, or when the equipment's operating status changes (such as switching from "normal operation" to "fault"), these changes may trigger new events, thereby adjusting the equipment's work schedule. During this process, trigger conditions can be dynamically set based on the equipment's real-time status, and these conditions provide the basis for subsequent event processing.

[0127] The workload of equipment refers to the intensity of work undertaken by a bucket wheel excavator or other equipment when performing a task. For example, the load or grasping capacity of a bucket wheel excavator during transportation. Changes in workload directly affect the operating efficiency and task execution time of the equipment. When the workload exceeds the equipment's capacity, it may cause changes in the equipment's operating status, trigger fault detection events, or require task reassignment. The system needs to adjust the sequence of events based on the real-time workload of the equipment to ensure that each piece of equipment can operate under an appropriate load and avoid overload.

[0128] Changes in the operating environment refer to external conditions that affect equipment operation, such as weather conditions and the accessibility of transportation routes. For example, severe weather conditions or traffic problems may reduce the operating efficiency of bucket wheel excavators or even delay task execution. Changes in the operating environment may trigger events that adjust the timing of task execution, such as changing transportation routes or adjusting task allocation, to adapt to new environmental conditions. By dynamically adjusting the timing of events, the system can respond to these external changes in a timely manner, ensuring the smooth execution of tasks.

[0129] A change in transport route refers to an adjustment to the originally planned transport route due to changes in equipment status or environmental factors. For example, in some cases, equipment failure or environmental impacts may force the system to replan the transport path. Adjustments to the transport route will trigger a series of events, such as adjustments to equipment scheduling and updates to work plans. By combining changes in transport routes with event triggering mechanisms, it can be ensured that the bucket wheel excavator always selects the most suitable path to complete the task and optimize overall operational efficiency.

[0130] Ultimately, the dynamic simulation scenario refers to the real-time simulation result generated through the aforementioned event triggering and timing adjustments. The dynamic simulation scenario can reflect the changes that may occur in the system during actual operation and adjust the equipment's operating status, task allocation, and transportation routes based on real-time input. This simulation scenario provides real-time data support for subsequent optimization and scheduling, ensuring that the system can respond quickly to real-time conditions and avoid equipment idleness, malfunctions, or inefficient operations.

[0131] Step S430: Efficiency evaluation is performed based on discrete event sequences and dynamic simulation scenarios. By constructing discrete event models, the transportation process under different operation schemes is simulated using the Monte Carlo simulation algorithm to obtain the efficiency simulation results of each scheme.

[0132] Understandably, the discrete event model transforms factors such as transportation tasks, equipment operating status, and working environment into discrete events, and combines these with the timing of these events to form a complete simulation model. This model can reflect the execution of various tasks during the operation of the power plant and bucket wheel excavator, as well as the possible state changes of the equipment during operation. Through the discrete event model, the system can perform detailed simulations of the execution process of different work plans and provide key data such as the timing information and task execution status of each event.

[0133] Next, the Monte Carlo simulation algorithm is applied to simulate the transportation process under different operational schemes. Monte Carlo simulation is a statistical simulation method based on random sampling, capable of simulating complex systems through a large number of random trials. In this step, the Monte Carlo simulation algorithm generates a large number of random event sequences based on a discrete event model and simulates the execution of the transportation process based on each sequence. These random trials consider factors such as changes in equipment status, allocation of transportation tasks, and fluctuations in the operating environment. Through repeated simulations, the efficiency distribution under different operational schemes can be obtained, and the performance of different operational schemes can be evaluated.

[0134] The specific simulation process includes the following aspects:

[0135] Random event simulation: Using the Monte Carlo simulation algorithm, multiple different event occurrence sequences are generated based on the probability distribution of discrete event sequences.

[0136] Implementation evaluation: Each event sequence corresponds to a different work plan. During the simulation, the work plan will be dynamically adjusted according to the equipment operating status and environmental changes to simulate the execution of the transportation task.

[0137] Efficiency index calculation: Through multiple simulations, the efficiency index (such as completion time, energy consumption, equipment utilization rate, etc.) of each work plan is calculated to obtain the overall performance of the plan.

[0138] The final efficiency simulation results are evaluations of different operational schemes, showing the efficiency of each scheme under various conditions. These results can be used to compare the advantages and disadvantages of different schemes, helping power plants select the optimal operational scheme. Through this simulation, the system can deeply analyze bottlenecks in the transportation process, optimize resource allocation, reduce time and energy waste during operations, thereby improving overall operational efficiency.

[0139] Further, step S500 includes steps S510 to S530.

[0140] Step S510: Set the objective function based on the efficiency simulation results. Quantify transportation efficiency, equipment failure rate, energy consumption, and environmental impact by analyzing the simulation results to obtain the quantitative indicators for optimizing the objective function.

[0141] It should be noted that several main optimization objectives need to be considered when setting the objective function:

[0142] Transportation efficiency: This objective primarily focuses on the efficiency of the material transportation process, typically quantified by indicators such as task completion time and the ratio of transportation volume to time. Preferably, transportation efficiency can be reflected by calculating the transportation volume or task completion time per unit of time.

[0143] Equipment failure rate: Equipment failure rate is an important indicator for measuring equipment reliability. It is usually calculated as the probability of equipment failure within a certain period of time. By analyzing the occurrence of equipment failures in simulation results, quantitative data on the equipment failure rate can be obtained, thus reflecting the overall stability of the system.

[0144] Energy consumption: Energy consumption mainly refers to the energy consumed during transportation, usually quantified by energy consumption per unit of transport volume or the total energy consumption of the entire system. Efficiency simulation results can be used to assess the impact of different operational plans on energy use, thereby optimizing energy utilization during operations.

[0145] Environmental impact: Environmental impact is a key indicator for evaluating the sustainability of an operational plan, typically including factors such as carbon emissions and noise. Simulation results can provide data on the environmental impact of different plans, thus helping to select the most environmentally friendly operational method.

[0146] Step S520: Perform weighted combination processing based on the quantitative indicators of the optimization objective function, determine the weight of each objective through the particle swarm optimization algorithm, and calculate the comprehensive efficiency of each operation plan to obtain the weighted objective value.

[0147] As is understandable, weighted combination processing combines different optimization objectives (such as transportation efficiency, equipment failure rate, energy consumption, and environmental impact) into a unified objective function. In multi-objective optimization problems, the relative importance of each objective may differ, thus requiring the assignment of a weight to each objective. The weighted combination process merges the quantitative indicators of each optimization objective into a unified objective function by assigning weights. The key to this process lies in how to allocate the weights of different objectives based on actual needs and priorities.

[0148] Particle Swarm Optimization (PSO) is a swarm intelligence-based optimization algorithm widely used in multi-objective optimization problems. In this step, the task of PSO is to determine the weights of each objective. PSO simulates the movement and search process of a swarm of particles to find the optimal weight combination that optimizes the overall objective function. Each particle represents a potential weight combination scheme. Particles move in the search space and adjust their positions according to the fitness function (i.e., the value of the overall objective function). Through continuous iteration, the particle swarm gradually converges to the globally optimal weight combination. PSO can automatically adjust the weights of each objective, achieving the optimal trade-off between objectives.

[0149] Step S530: Solve the weighted objective value, use the simulated annealing algorithm to find the optimal solution in the multi-objective space, and generate a specific operation plan based on the optimal solution.

[0150] It's important to note that, firstly, simulated annealing is a heuristic global optimization algorithm inspired by the physical annealing process. Simulated annealing simulates the process of a substance decreasing in temperature from high to low, gradually reducing the system's "temperature" during the search, thus guiding the search from exploring the global optimum to gradually converging towards a local optimum. For multi-objective optimization problems, simulated annealing effectively avoids getting trapped in local optima by using random perturbations and acceptance probability strategies to perform a global search and find the optimal trade-off between multiple objectives.

[0151] A multi-objective space refers to a high-dimensional space formed after considering multiple optimization objectives such as transportation efficiency, equipment failure rate, energy consumption, and environmental impact. In this space, each point represents a possible operational plan, containing the weighted values ​​of each objective and corresponding quantitative indicators. The simulated annealing algorithm searches for the optimal solution by iterating within this multi-dimensional space. The algorithm gradually approaches the optimal solution by accepting better solutions and, with a certain probability, accepting worse solutions (to avoid getting trapped in local optima).

[0152] In the specific solution process, the simulated annealing algorithm is optimized through the following steps:

[0153] Step S531, Initial Solution Selection: Starting from a random initial job scheme, calculate its weighted objective value (overall efficiency).

[0154] Step S532, Neighborhood Search: Generate new neighborhood solutions by perturbing the current solution (i.e., the work plan). For example, adjust the allocation of equipment tasks and the selection of transportation routes.

[0155] Step S533, Acceptance Criterion: By calculating the weighted objective value of the new solution, decide whether to accept the solution. If the overall performance of the new solution is better than the current solution, then accept the new solution; if the new solution is worse, then accept the solution with a certain probability, thereby avoiding getting trapped in a local optimum.

[0156] Step S534, Temperature Attenuation: As the algorithm iterates, the "temperature" is gradually reduced to decrease the probability of accepting poor solutions, thereby making the search process more focused on local optima.

[0157] Step S535, Termination Condition: When the predetermined number of iterations is reached or the temperature drops to a set threshold, the algorithm stops and outputs the final optimal solution.

[0158] The optimal solution represents the best trade-off among various optimization objectives within a multi-objective space. This optimal solution ensures that the power plant achieves an optimal balance in terms of transportation efficiency, equipment failure rate, energy consumption, and environmental impact, thus meeting the overall optimization requirements of the system.

[0159] Finally, the specific operational plan is generated based on the optimal solution to formulate the actual work task arrangement. For example, the transport volume of each bucket wheel excavator, the operating path of the equipment, and the scheduling of operating periods are adjusted according to the optimal solution. This operational plan will serve as a guide for the power plant in actual operation, ensuring the optimal performance of the system under multiple objectives.

[0160] Example 2:

[0161] like Figure 2 As shown, this embodiment provides an unmanned intelligent operation system for power plant bucket wheel excavators based on AI algorithms. The system includes:

[0162] The acquisition module 901 is used to acquire power plant operating status information and bucket wheel excavator information. The power plant status information includes fuel reserves, boiler load, turbine status and power generation. The bucket wheel excavator information includes the bucket wheel excavator's location information, transportation efficiency, material grabbing volume and equipment fault status.

[0163] Modeling module 902 is used to construct a digital twin model of power plant-bucket wheel machine by performing trend analysis and state change modeling on the equipment status of power plant and the operating status of bucket wheel machine based on the power plant's operating status information and bucket wheel machine information.

[0164] Analysis module 903 performs correlation analysis between power plant fuel reserves and bucket wheel excavator transportation tasks based on a digital twin model of power plant-bucket wheel excavator. It simulates transportation routes through path planning algorithms and evaluates loading and unloading efficiency and equipment coordination by combining real-time status data to obtain an optimized transportation plan.

[0165] The simulation module 904 is used to perform discrete event simulation processing based on the optimized transportation scheme. By converting the transportation process into discrete events and combining real-time equipment status data for data format conversion and event-driven simulation, efficiency simulation results are obtained.

[0166] The optimization module 905 performs optimization based on the efficiency simulation results. By setting objective functions, including maximizing transportation efficiency, minimizing equipment failure rate, optimizing energy consumption and minimizing environmental impact, a multi-objective optimization algorithm is used to weight and combine the objectives, and the optimal solution is obtained to obtain the operation plan.

[0167] In some embodiments disclosed in this application, the modeling module 902 includes:

[0168] The first modeling unit is used to perform equipment status trend analysis based on the power plant's operating status information. By using the autoregressive integral moving average algorithm, it performs trend analysis on the power plant's fuel reserves, boiler load, turbine status, and power generation to obtain future predicted values ​​of the power plant's equipment status.

[0169] The second modeling unit is used to model the state changes based on the operating state information of the bucket wheel excavator. It models the state changes of the bucket wheel excavator by constructing a hidden Markov model and simulates the operating state changes of the bucket wheel excavator to obtain the state transition probability matrix of the bucket wheel excavator.

[0170] The third modeling unit is used to acquire historical data on the status of power plant equipment and the operating status of bucket wheel excavators, and to perform synergy effect analysis based on historical data. It also performs association rule mining analysis on the correlation between the status of power plant equipment and the status of bucket wheel excavators to obtain the equipment collaborative working mode.

[0171] The fourth modeling unit is used to extract features from the power plant and bucket wheel machine data of various dimensions using a convolutional neural network based on the future predicted values ​​of the power plant equipment status, the state transition probability matrix of the bucket wheel machine, and the equipment collaborative working mode, and to use an ensemble learning algorithm to fuse the models and construct a digital twin model of the power plant and bucket wheel machine.

[0172] In some embodiments disclosed in this application, the second modeling unit includes:

[0173] The fifth modeling unit is used to perform state extraction processing based on the operating status information of the bucket wheel excavator. By using a state switching point detection algorithm to analyze the historical operating data of the bucket wheel excavator, the switching point sequence of each state of the bucket wheel excavator is obtained.

[0174] The sixth modeling unit is used to perform state transition modeling based on the switching point sequence. It models the state change process of the bucket wheel machine by constructing a hidden Markov model and uses the maximum likelihood estimation algorithm to train the model and obtain the probabilistic framework of state transition.

[0175] The seventh modeling unit performs approximate probability calculations based on the probabilistic framework of state transitions. By statistically analyzing and normalizing the transition frequencies between states, the state transition probability matrix of the bucket wheel machine is obtained.

[0176] Example 3:

[0177] Corresponding to the above method embodiments, this embodiment also provides an unmanned intelligent operation device for power plant bucket wheel excavators based on AI algorithms. The unmanned intelligent operation device for power plant bucket wheel excavators based on AI algorithms described below and the unmanned intelligent operation method for power plant bucket wheel excavators based on AI algorithms described above can be referred to in correspondence.

[0178] Figure 3 This is a block diagram illustrating an AI-based unmanned intelligent operation device 800 for a power plant bucket wheel excavator, according to an exemplary embodiment. Figure 3 As shown, the AI-based unmanned intelligent operation equipment 800 for power plant bucket wheel excavators may include: a processor 801 and a memory 802. The AI-based unmanned intelligent operation equipment 800 may also include one or more of the following: a multimedia component 803, an I / O interface 804, and a communication component 805.

[0179] The processor 801 controls the overall operation of the AI-based unmanned intelligent power plant bucket wheel excavator 800 to complete all or part of the steps in the aforementioned AI-based unmanned intelligent power plant bucket wheel excavator operation method. The memory 802 stores various types of data to support the operation of the AI-based unmanned intelligent power plant bucket wheel excavator 800. This data may include, for example, instructions for any application or method operating on the AI-based unmanned intelligent power plant bucket wheel excavator 800, as well as application-related data such as contact data, sent and received messages, images, audio, video, etc. The memory 802 can be implemented using any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touchscreen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in the memory 802 or transmitted via the communication component 805. The audio component also includes at least one speaker for outputting audio signals. I / O interface 804 provides an interface between processor 801 and other interface modules, such as keyboards, mice, and buttons. These buttons can be virtual or physical. Communication component 805 is used for wired or wireless communication between the AI-based unmanned intelligent power plant bucket wheel excavator 800 and other devices. Wireless communication methods include Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, or 4G, or a combination thereof. Therefore, the corresponding communication component 805 may include a Wi-Fi module, a Bluetooth module, and an NFC module.

[0180] In an exemplary embodiment, an AI-based unmanned intelligent operation device 800 for a power plant bucket wheel excavator can be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to execute the aforementioned AI-based unmanned intelligent operation method for a power plant bucket wheel excavator.

[0181] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided. When executed by a processor, these program instructions implement the steps of the aforementioned AI-based unmanned intelligent operation method for a power plant bucket wheel excavator. For example, the computer-readable storage medium may be the aforementioned memory 802 including program instructions, which may be executed by a processor 801 of an AI-based unmanned intelligent operation device 800 for a power plant bucket wheel excavator to complete the aforementioned AI-based unmanned intelligent operation method for a power plant bucket wheel excavator.

[0182] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. An AI algorithm-based unmanned intelligent operation method for a bucket wheel machine in a power plant, characterized by, The method comprises the following steps: acquiring power plant operation state information and bucket wheel machine information, wherein the power plant state information comprises fuel reserves, boiler load, turbine state and power generation power, and the bucket wheel machine information comprises position information, transportation efficiency, material grabbing amount and equipment fault state of the bucket wheel machine; based on the power plant operation state information and the bucket wheel machine information, a power plant-bucket wheel machine digital twin model is constructed by performing trend analysis and state change modeling on the equipment state of the power plant and the operation state of the bucket wheel machine; based on the power plant-bucket wheel machine digital twin model, correlation analysis is performed on the fuel reserves of the power plant and the transportation task of the bucket wheel machine, a transportation route is simulated by using a path planning algorithm, and the loading and unloading efficiency and the equipment collaborative working condition are evaluated in combination with real-time state data to obtain an optimized transportation scheme; according to the optimized transportation scheme, discrete event simulation processing is performed, the transportation process is converted into discrete events, data format conversion and event-driven simulation are performed in combination with real-time equipment state data, and efficiency simulation results are obtained; based on the efficiency simulation results, optimization processing is performed, a target function is set, the target function comprises maximizing transportation efficiency, minimizing equipment failure rate, optimizing energy consumption and minimizing environmental impact, a multi-objective optimization algorithm is used to weight and combine each target, and an optimal solution is solved to obtain a work scheme. 2.The AI algorithm-based unmanned intelligent operation method of a power plant bucket wheel machine according to claim 1, characterized in that, According to the power plant operation state information and the bucket wheel machine information, a power plant-bucket wheel machine digital twin model is constructed by performing trend analysis and state change modeling on the equipment state of the power plant and the operation state of the bucket wheel machine, comprising: performing equipment state trend analysis according to the power plant operation state information, performing trend analysis on the fuel reserves, boiler load, turbine state and power generation power of the power plant by using an autoregressive integrated moving average algorithm, and obtaining future predicted values of the power plant equipment state; performing state change modeling according to the bucket wheel machine operation state information, modeling the state change of the bucket wheel machine by constructing a hidden Markov model, simulating the operation state change of the bucket wheel machine, and obtaining a bucket wheel machine state transition probability matrix; acquiring historical data of the power plant equipment state and the bucket wheel machine operation state, and performing collaborative effect analysis based on the historical data, and mining correlation rules to analyze the correlation between the power plant equipment state and the bucket wheel machine state to obtain a device collaborative working mode; according to the future predicted values of the power plant equipment state, the bucket wheel machine state transition probability matrix and the device collaborative working mode, a convolutional neural network is used to extract features of each dimension of data of the power plant and the bucket wheel machine, and an integrated learning algorithm is used for model fusion to construct a power plant-bucket wheel machine digital twin model. 3.The AI algorithm-based intelligent operation method of the reclaimer of the power plant without human intervention according to claim 2, characterized in that, According to the bucket wheel machine operation state information, a hidden Markov model is constructed to model the state change of the bucket wheel machine, the operation state change of the bucket wheel machine is simulated, and a bucket wheel machine state transition probability matrix is obtained, comprising: performing state extraction processing according to the bucket wheel machine operation state information, analyzing historical operation data of the bucket wheel machine by using a state switching point detection algorithm to obtain a switching point sequence of each state of the bucket wheel machine; According to the state transition modeling process of the switching point sequence, the state change process of the bucket wheel machine is modeled by constructing a hidden Markov model, and the model is trained using a maximum likelihood estimation algorithm to obtain a state transition probability framework; According to the state transition probability framework, a rough probability calculation process is performed, the transition frequency between states is counted and normalized to obtain a bucket wheel machine state transition probability matrix. 4.The AI algorithm-based intelligent operation method of an unmanned reclaimer in a power plant according to claim 2, characterized in that, According to the future predicted value of the power plant equipment state, the bucket wheel machine state transition probability matrix and the equipment collaborative working mode, a convolutional neural network is used to extract features from each dimension of the data of the power plant and the bucket wheel machine, and an integrated learning algorithm is used for model fusion to construct a power plant-bucket wheel machine digital twin model, including: According to the future predicted value of the power plant equipment state, a physical process modeling is performed, the energy flow and efficiency change of the equipment in the future time period are simulated by combining the thermodynamic model and the energy balance equation, and the thermodynamic dynamic change prediction result of the power plant equipment is obtained; According to the state transition probability matrix and the equipment collaborative working mode, a modeling process is performed, the motion state and mechanical synergy effect of the bucket wheel machine are simulated by mechanical modeling and mechanical simulation, and the mechanical motion state simulation result of the bucket wheel machine is obtained; According to the thermodynamic dynamic change prediction result and the mechanical motion state simulation result, feature extraction is performed, and spatiotemporal information is extracted from the equipment state, energy efficiency and mechanical motion path dimensions of the power plant equipment and the bucket wheel machine by a convolutional neural network to obtain a feature vector; According to the feature vector, a model fusion and optimization process is performed, the output results of the thermodynamic model and the mechanical motion model are optimized by using an integrated learning algorithm, and a power plant-bucket wheel machine digital twin model is obtained. 5.The AI algorithm-based intelligent operation method of an unmanned reclaimer in a power plant according to claim 1, characterized in that, Based on the power plant-bucket wheel machine digital twin model, the correlation analysis of the fuel reserve quantity of the power plant and the transportation task of the bucket wheel machine is performed, the transportation route is simulated by a path planning algorithm, and the loading and unloading efficiency and the equipment collaborative working condition are evaluated in combination with real-time state data to obtain an optimized transportation scheme, including: According to the equipment state information in the power plant-bucket wheel machine digital twin model, a transportation task demand prediction process is performed, a long short-term memory network algorithm is used to analyze the fuel reserve quantity and the boiler load change trend to obtain a predicted transportation demand; According to the transportation demand, a task allocation process is performed in combination with the bucket wheel machine operation capacity, the task quantity of each bucket wheel machine is allocated by a mixed integer programming algorithm to obtain a specific operation plan of the bucket wheel machine; According to the specific operation plan, a path planning process is performed, the transportation route of the bucket wheel machine is simulated by a dynamic path planning model, and the optimal transportation path is obtained in combination with the equipment state change and the environmental conditions; According to the optimal transportation path, a loading and unloading efficiency evaluation and equipment collaborative working process is performed, the efficiency of different paths and collaborative operation schemes is analyzed to obtain an optimized transportation scheme. 6.The AI algorithm-based intelligent operation method for unmanned reclaimer of a power plant according to claim 1, characterized in that, According to the optimized transportation scheme, a discrete event simulation process is performed, the transportation process is converted into discrete events, and data format conversion and event-driven simulation are performed in combination with real-time equipment state data to obtain an efficiency simulation result, including: According to the optimized transportation scheme, event definition processing is performed, discrete events are obtained by converting the transportation task, loading and unloading process and equipment running state into discrete events and assigning time stamps, and a discrete event sequence is obtained; According to the power plant-bucket wheel digital twin model, event triggering processing is performed, the event occurrence timing is dynamically adjusted by combining the changes in the work load of the equipment, the working environment and the transportation route, and a dynamic simulation scenario is obtained; According to the discrete event sequence and the dynamic simulation scenario, efficiency evaluation processing is performed, a discrete event model is constructed, the transportation process under different operation schemes is simulated through a Monte Carlo simulation algorithm, and efficiency simulation results of each scheme are obtained. 7.The AI algorithm-based intelligent operation method of an unmanned reclaimer in a power plant according to claim 1, characterized in that, Based on the efficiency simulation results, optimization processing is performed, a target function is set, the target function includes maximizing transportation efficiency, minimizing equipment failure rate, optimizing energy consumption and minimizing environmental impact, a multi-objective optimization algorithm is used to weight and combine each target, and an optimal solution is solved to obtain an operation scheme, including: According to the efficiency simulation results, a target function is set, the transportation efficiency, equipment failure rate, energy consumption and environmental impact are quantified by analyzing the simulation results, and the quantitative indicators of the optimization target function are obtained; According to the quantitative indicators of the optimization target function, weighting combination processing is performed, the weights of each target are determined through a particle swarm optimization algorithm, the comprehensive performance of each operation scheme is calculated, and a weighted target value is obtained; Based on the weighted target value, solving processing is performed, the simulated annealing algorithm is used to find the optimal solution in the multi-objective space, and a specific operation scheme is generated based on the optimal solution.

8. An AI algorithm-based unmanned intelligent operation system for a bucket wheel machine in a power plant, characterized in that, It includes: An acquisition module is configured to acquire power plant operating state information and bucket wheel information, the power plant state information includes fuel reserves, boiler load, turbine state and power generation power, and the bucket wheel information includes position information, transportation efficiency, material grabbing amount and equipment failure state of the bucket wheel; A modeling module is configured to construct a power plant-bucket wheel digital twin model by performing trend analysis and state change modeling on the equipment state of the power plant and the running state of the bucket wheel based on the power plant operating state information and the bucket wheel information; An analysis module is configured to perform correlation analysis on the fuel reserves of the power plant and the transportation task of the bucket wheel based on the power plant-bucket wheel digital twin model, simulate the transportation route through a path planning algorithm, and evaluate the loading and unloading efficiency and the equipment cooperative working condition in combination with real-time state data to obtain an optimized transportation scheme; A simulation module is configured to perform discrete event simulation processing according to the optimized transportation scheme, convert the transportation process into discrete events, and perform data format conversion and event-driven simulation in combination with real-time equipment state data to obtain efficiency simulation results; An optimization module is configured to perform optimization processing based on the efficiency simulation results, set a target function, the target function includes maximizing transportation efficiency, minimizing equipment failure rate, optimizing energy consumption and minimizing environmental impact, use a multi-objective optimization algorithm to weight and combine each target, and solve an optimal solution to obtain an operation scheme. 9.The AI algorithm-based unmanned intelligent operation system for a bucket wheel machine in a power plant according to claim 8, characterized in that, The modeling module includes: The first modeling unit is configured to perform equipment state trend analysis according to the power plant operation state information, to obtain future predicted values of the power plant equipment state by performing trend analysis on fuel reserves, boiler load, steam turbine state and power generation capacity of the power plant using an autoregressive integrated moving average algorithm; The second modeling unit is configured to perform state change modeling according to the bucket wheel machine operation state information, to obtain a state transition probability matrix of the bucket wheel machine by modeling state changes of the bucket wheel machine using a hidden Markov model and simulating operation state changes of the bucket wheel machine; The third modeling unit is configured to obtain historical data of the power plant equipment state and the bucket wheel machine operation state, and to perform collaborative effect analysis based on the historical data, to mine association rules to analyze the correlation between the power plant equipment state and the bucket wheel machine state, and to obtain an equipment collaborative working mode; The fourth modeling unit is configured to perform feature extraction on each dimension of data of the power plant and the bucket wheel machine using a convolutional neural network based on the future predicted values of the power plant equipment state, the state transition probability matrix of the bucket wheel machine and the equipment collaborative working mode, and to perform model fusion using an ensemble learning algorithm to construct a power plant-bucket wheel machine digital twin model. 10.The AI algorithm-based unmanned intelligent operation system for a bucket wheel machine in a power plant according to claim 9, wherein, The second modeling unit includes: The fifth modeling unit is configured to perform state extraction processing according to the bucket wheel machine operation state information, to obtain a switching point sequence of each state of the bucket wheel machine by analyzing historical operation data of the bucket wheel machine using a state switching point detection algorithm; The sixth modeling unit is configured to perform state transition modeling processing according to the switching point sequence, to model a state change process of the bucket wheel machine by constructing a hidden Markov model, and to train the model using a maximum likelihood estimation algorithm to obtain a probability framework of state transition; The seventh modeling unit is configured to perform rough probability calculation processing based on the probability framework of state transition, to obtain a state transition probability matrix of the bucket wheel machine by statistically analyzing transition frequencies between states and performing normalization.

Images