Bucket wheel machine automatic control method and system based on deep belief network
Through the deep confidence network and improved Firefly algorithm, the PID controller parameters are optimized, and the problems of insufficient efficiency and lack of intelligence of the automatic control system of the bucket turbine are solved, efficient and safe automatic control is achieved, and system stability and production efficiency are improved.
Patent Information
- Application Number
- CN202410838574.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2024-05-20
- Filing Date
- 2024-06-26
- Publication Date
- 2024-11-12
AI Technical Summary
The existing bucket turbine automatic control system is insufficient in efficiency and lacks intelligent and real-time fault detection, resulting in high maintenance costs and safety hazards. The PLC control program optimization design space is large, affecting the accuracy and efficiency of automatic control.
Deep Confidence Network (DBN) combined with improved Firefly algorithm is adopted to identify the operating mode of the bucket turbine through unsupervised learning, optimize the PID controller parameters, realize automatic control and intelligent management, use PLC to collect real-time data for closed-loop control, reduce manual intervention and Improve system stability.
It improves the operating efficiency and safety of the bucket turbine, reduces maintenance costs, improves the accuracy of automatic control and system reliability, simplifies the algorithm structure and optimizes iterative steps, and realizes automated control of modern power plant operations.
Smart Images

Figure CN118938644A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of automatic control technology for bucket wheel excavators based on deep belief networks, and in particular to an automatic control method and system for bucket wheel excavators based on deep belief networks. Background Technology
[0002] After decades of development, the working capacity of domestic bucket wheel excavators has narrowed dramatically compared to foreign equipment. Research on bucket wheel excavators in my country has also deepened, particularly in intelligent control. Currently, the electrical control methods for domestic bucket wheel excavators mostly utilize single-machine control via PLC (Programmable Logic Controller), with some simple manual control operations. Domestic bucket wheel excavator control systems have gone through three development stages: local manual control, semi-automatic control, and automatic control. Communication between the bucket wheel excavator and the control center is achieved through relay contacts and traditional multi-core cables, while the control system monitors the equipment intelligently through traditional analog instrument panels. The operating status and parameters of field equipment in the main control room cannot be obtained in real time, and equipment alarm information is lacking, thus requiring manual information transmission. This control system suffers from low automation, poor operability, large size, complex wiring, and poor maintainability. Most importantly, it cannot achieve automatic management of equipment operating conditions.
[0003] Automatic control refers to the process of monitoring the operation of a bucket wheel excavator through a monitoring system, ultimately achieving unmanned on-site operation and remote control. From a remote monitoring room, staff can monitor the bucket wheel excavator's operation, including graphical representations of its working status, working position, fault locations, and automatic alarm systems. They can also control processes such as material handling, thus achieving automatic control. Staff can use computers to access the bucket wheel excavator's operating parameters and exchange data with the main control room to adjust parameters in real time. However, manual operation of bucket wheel excavators suffers from excessive subjectivity in coal pile placement planning and allocation, as well as unclear boundary conditions in the selection of feeding and retrieving areas. Furthermore, inconsistent operator proficiency can lead to errors that accelerate wear and tear on mechanical and control equipment, significantly increasing maintenance costs. Moreover, current bucket wheel excavators lack early fault diagnosis and real-time fault detection feedback, posing significant safety hazards to personnel. While the automatic control system for bucket wheel excavators has solved the problems associated with manual operation, eliminating the need for highly skilled operators and delegating more complex tasks to computers, and avoiding numerous safety hazards, it still suffers from inefficiencies after decades of development. In particular, for existing programmable logic controllers (PLCs), the optimization of the program design has a significant impact on the accuracy of automatic control and the efficiency of the control system.
[0004] The optimization design of existing programmable logic controller (PLC) control programs has a significant impact on the accuracy of automatic control and the efficiency of the control system, indicating that there may be room for improvement in program design. Bucket wheel excavators on the market lack early fault judgment feedback and real-time detection feedback, which may lead to untimely fault detection, increasing maintenance costs and safety risks. Summary of the Invention
[0005] This application provides an automatic control method and system for bucket wheel excavators based on deep belief networks, aiming to address the shortcomings of insufficient efficiency in existing technologies.
[0006] Firstly, an automatic control method for bucket wheel excavators based on deep belief networks, the method comprising:
[0007] The operating data of the bucket wheel excavator is collected using a PLC controller, and the operating data includes the input and output parameters of the PLC controller. The collected operating data is used as the input of a DBN (Database Network), and the network is trained through unsupervised learning so that it can identify the patterns and trends in the operation of the bucket wheel excavator.
[0008] Based on the initial operating conditions of the bucket wheel excavator, set the initial parameters for the PID controller;
[0009] The attributes of the firefly population are defined based on the firefly algorithm, and the position of each firefly is calculated through an objective function. The optimal solution of the PID parameters is found by simulating the firefly's bioluminescence behavior. In each iteration of the firefly population, the iteration factor is adjusted according to the current performance of the algorithm to adaptively adjust the search range. If the firefly population does not evolve in 5 consecutive iterations, Gaussian mutation is introduced, and the positions of the fireflies are updated according to a specific mutation strategy.
[0010] Based on the optimized Firefly algorithm results, the parameters of the PID controller are updated, and the updated PID parameters are written into the PLC to realize the automatic control of the bucket wheel excavator.
[0011] Optionally, in the above scheme, the step of using a PLC controller to collect the operating data of the bucket wheel excavator includes the following operating data: (The PLC controller's input and output parameters are listed below.)
[0012] The PLC system is used to collect real-time operating data of the bucket wheel excavator, including motor current, voltage, temperature, and vibration frequency.
[0013] Provide accurate input data for DBN training to ensure that the control strategy can make decisions based on actual operating conditions.
[0014] In the above scheme, optionally, the initial parameters of the PID controller are set including: setting the proportional coefficient, integral time, and derivative time.
[0015] Optionally, in the above scheme, updating the parameters of the PID controller based on the optimized firefly algorithm results includes:
[0016] The optimized PID parameters are written into the PLC to replace the original parameter settings. The control strategy of the bucket wheel excavator is adjusted according to the optimization results to improve control accuracy and efficiency.
[0017] Optionally, in the above scheme, writing the updated PID parameters into the PLC to achieve automatic control of the bucket wheel excavator includes:
[0018] By utilizing the updated PID parameters, automatic control of the bucket wheel excavator is achieved through a PLC, realizing intelligent control of the bucket wheel excavator and improving production efficiency and safety.
[0019] Optionally, in the above scheme, after writing the updated PID parameters into the PLC to achieve automatic control of the bucket wheel excavator, the method further includes:
[0020] The system continuously monitors the operating status of the bucket wheel excavator and feeds real-time data back to the DBN network and the Firefly algorithm to achieve closed-loop control. The control strategy is adjusted based on the real-time data to ensure that the bucket wheel excavator is always in the optimal operating state.
[0021] In the above scheme, optionally, after writing the updated PID parameters into the PLC to realize the automatic control of the bucket wheel excavator, it also includes: regularly maintaining the system, checking the performance of the DBN network and the firefly algorithm, and making adjustments and optimizations as needed to ensure the long-term stable operation of the system.
[0022] Secondly, an automatic control system for a bucket wheel excavator based on a deep belief network, the system comprising:
[0023] Data Acquisition Module: Used to collect operating data of the bucket wheel excavator using a PLC controller, the operating data including the input and output parameters of the PLC controller; the collected operating data is used as the input of the DBN, and the network is trained through unsupervised learning so that it can identify the patterns and trends in the operation of the bucket wheel excavator;
[0024] Initialization module: Used to set initial parameters for the PID controller based on the initial operating conditions of the bucket wheel excavator;
[0025] The optimization module is used to define the attributes of the firefly population based on the firefly algorithm, calculate the position of each firefly through the objective function, and find the optimal solution of the PID parameters by simulating the firefly's bioluminescence behavior. In each iteration of the firefly population, the iteration factor is adjusted according to the current performance of the algorithm to adaptively adjust the search range. If the firefly population has not evolved in 5 consecutive iterations, Gaussian mutation is introduced to update the firefly positions according to a specific mutation strategy.
[0026] Update module: Used to update the parameters of the PID controller based on the optimized Firefly algorithm results, and write the updated PID parameters into the PLC to realize the automatic control of the bucket wheel excavator.
[0027] Thirdly, a computer device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to perform the following steps:
[0028] The operating data of the bucket wheel excavator is collected using a PLC controller, and the operating data includes the input and output parameters of the PLC controller. The collected operating data is used as the input of a DBN (Database Network), and the network is trained through unsupervised learning so that it can identify the patterns and trends in the operation of the bucket wheel excavator.
[0029] Based on the initial operating conditions of the bucket wheel excavator, set the initial parameters for the PID controller;
[0030] The attributes of the firefly population are defined based on the firefly algorithm, and the position of each firefly is calculated through an objective function. The optimal solution of the PID parameters is found by simulating the firefly's bioluminescence behavior. In each iteration of the firefly population, the iteration factor is adjusted according to the current performance of the algorithm to adaptively adjust the search range. If the firefly population does not evolve in 5 consecutive iterations, Gaussian mutation is introduced, and the positions of the fireflies are updated according to a specific mutation strategy.
[0031] Based on the optimized Firefly algorithm results, the parameters of the PID controller are updated, and the updated PID parameters are written into the PLC to realize the automatic control of the bucket wheel excavator.
[0032] Fourthly, a computer program product includes a computer program / instructions that, when executed by a processor, perform the following steps:
[0033] The operating data of the bucket wheel excavator is collected using a PLC controller, and the operating data includes the input and output parameters of the PLC controller. The collected operating data is used as the input of a DBN (Database Network), and the network is trained through unsupervised learning so that it can identify the patterns and trends in the operation of the bucket wheel excavator.
[0034] Based on the initial operating conditions of the bucket wheel excavator, set the initial parameters for the PID controller;
[0035] The attributes of the firefly population are defined based on the firefly algorithm, and the position of each firefly is calculated through an objective function. The optimal solution of the PID parameters is found by simulating the firefly's bioluminescence behavior. In each iteration of the firefly population, the iteration factor is adjusted according to the current performance of the algorithm to adaptively adjust the search range. If the firefly population does not evolve in 5 consecutive iterations, Gaussian mutation is introduced, and the positions of the fireflies are updated according to a specific mutation strategy.
[0036] Based on the optimized Firefly algorithm results, the parameters of the PID controller are updated, and the updated PID parameters are written into the PLC to realize the automatic control of the bucket wheel excavator.
[0037] Compared with the prior art, this application has at least the following beneficial effects:
[0038] Based on further analysis and research of existing technical problems, this application recognizes that for existing programmable logic controller (PLC) control, the optimized design of the program has a significant impact on the accuracy of automatic control and the efficiency of the control system. Therefore, optimizing the control algorithm used for calculation is essential. This application utilizes a PLC controller to collect operating data of the bucket wheel excavator, including the PLC controller's input and output parameters. The collected operating data is used as input to a deep learning network (DBN), which is trained through unsupervised learning to identify the operating patterns and trends of the bucket wheel excavator. Based on the initial operating conditions of the bucket wheel excavator, a PID controller is then developed. The controller sets initial parameters; attributes of the firefly population are defined based on the firefly algorithm, and the position of each firefly is calculated through an objective function. The optimal solution for the PID parameters is found by simulating the firefly's bioluminescence behavior. In each iteration of the firefly population, the iteration factor is adjusted according to the current performance of the algorithm to adaptively adjust the search range. If the firefly population does not evolve in 5 consecutive iterations, Gaussian mutation is introduced, and the firefly positions are updated according to a specific mutation strategy. Based on the optimized firefly algorithm results, the parameters of the PID controller are updated, and the updated PID parameters are written into the PLC to realize the automatic control of the bucket wheel excavator.
[0039] This solution, by combining deep belief networks and an improved firefly algorithm, provides a solution to the problems existing in current technologies. Through automated and intelligent control strategies, it can improve the operating efficiency and safety of bucket wheel excavators, reduce manual intervention, lower maintenance costs, and enhance system stability and reliability.
[0040] The automatic control strategy proposed in this invention can greatly improve operational efficiency, improve the working environment of the workforce and ensure the safety of workers, enhance the level of automatic control of power plant bucket wheel excavators, and further realize the modernization of power plant operations. Compared with existing automatic control schemes, this invention simplifies the algorithm structure and optimizes the iteration steps, which is of great significance for the research on optimization algorithms of programmable controllers. Attached Figure Description
[0041] Figure 1 A flowchart illustrating an automatic control method for a bucket wheel excavator based on a deep belief network, provided as an embodiment of this application;
[0042] Figure 2 A schematic diagram of the control principle of a closed-loop speed regulation system for an asynchronous motor provided in one embodiment of this application;
[0043] Figure 3 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation
[0044] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.
[0045] In one embodiment, such as Figure 1 and Figure 2 As shown, an automatic control method for bucket wheel excavators based on deep belief networks is provided, including the following steps:
[0046] The operating data of the bucket wheel excavator is collected using a PLC controller, and the operating data includes the input and output parameters of the PLC controller. The collected operating data is used as the input of a DBN (Database Network), and the network is trained through unsupervised learning so that it can identify the patterns and trends in the operation of the bucket wheel excavator.
[0047] Based on the initial operating conditions of the bucket wheel excavator, set the initial parameters for the PID controller;
[0048] The attributes of the firefly population are defined based on the firefly algorithm, and the position of each firefly is calculated through an objective function. The optimal solution of the PID parameters is found by simulating the firefly's bioluminescence behavior. In each iteration of the firefly population, the iteration factor is adjusted according to the current performance of the algorithm to adaptively adjust the search range. If the firefly population does not evolve in 5 consecutive iterations, Gaussian mutation is introduced, and the positions of the fireflies are updated according to a specific mutation strategy.
[0049] Based on the optimized Firefly algorithm results, the parameters of the PID controller are updated, and the updated PID parameters are written into the PLC to realize the automatic control of the bucket wheel excavator.
[0050] In this embodiment, the operation data of the bucket wheel excavator is collected using a PLC controller. This operation data includes the input and output parameters of the PLC controller, including:
[0051] The PLC system is used to collect real-time operating data of the bucket wheel excavator, including but not limited to motor current, voltage, temperature, and vibration frequency.
[0052] Provide accurate input data for DBN training to ensure that the control strategy can make decisions based on actual operating conditions.
[0053] In this embodiment, the initial parameters set for the PID controller specifically include: setting the proportional coefficient, integral time, and derivative time.
[0054] In this embodiment, updating the parameters of the PID controller based on the optimized firefly algorithm results includes:
[0055] The optimized PID parameters are written into the PLC to replace the original parameter settings. The control strategy of the bucket wheel excavator is adjusted according to the optimization results to improve control accuracy and efficiency.
[0056] In this embodiment, writing the updated PID parameters into the PLC to achieve automatic control of the bucket wheel excavator includes:
[0057] By utilizing the updated PID parameters, automatic control of the bucket wheel excavator is achieved through a PLC, realizing intelligent control of the bucket wheel excavator and improving production efficiency and safety.
[0058] In this embodiment, after writing the updated PID parameters into the PLC to achieve automatic control of the bucket wheel excavator, the method further includes:
[0059] The system continuously monitors the operating status of the bucket wheel excavator and feeds real-time data back to the DBN network and the Firefly algorithm to achieve closed-loop control. The control strategy is adjusted based on the real-time data to ensure that the bucket wheel excavator is always in the optimal operating state.
[0060] In this embodiment, after writing the updated PID parameters into the PLC to realize the automatic control of the bucket wheel excavator, the system also includes: regularly maintaining the system, checking the performance of the DBN network and the Firefly algorithm, and making adjustments and optimizations as needed to ensure long-term stable operation of the system.
[0061] The aforementioned automatic control method for bucket wheel excavators based on deep belief networks combines deep belief networks with an improved firefly algorithm to provide a solution to existing problems. Through automated and intelligent control strategies, the operating efficiency and safety of the bucket wheel excavator can be improved, manual intervention reduced, maintenance costs lowered, and system stability and reliability enhanced.
[0062] In one embodiment, such as Figure 2 As shown, to improve the automation control level of power plants, belonging to the field of automatic control, this paper studies the control strategy of PLC-based bucket wheel excavators, and investigates the application optimization of deep belief networks (DBNs). A proportional-integral-control (PID) parameter update strategy based on DBNs is proposed to achieve intelligent control of the bucket wheel excavator. This invention, based on the nonlinear fitting capability of DBNs, collects PLC inputs and outputs to complete network training. Simultaneously, to improve the efficiency of DBN network training, an improved firefly algorithm is introduced when updating PID parameters, ensuring iterative stability and avoiding stagnation.
[0063] Under the guidance of intelligent control strategies, the bucket wheel excavator will automatically execute material handling processes such as coal feeding and discharging according to software instructions, and monitor and collect information on materials and equipment in real time. Operators can modify the system's operation process at any time from a remote monitoring console according to actual work needs, making the unmanned operation of the bucket wheel excavator more closely aligned with production process requirements and maximizing its production value. This invention utilizes the production data collected by the bucket wheel excavator to drive the intelligent control strategy, and then writes deep learning and intelligent optimization algorithms into the PLC; ultimately, it realizes parameter updates in the proportional-integral-control (PID) controller in the PLC, achieving automatic control through parameter updates.
[0064] With the rapid development of computer hardware, artificial intelligence algorithms have quickly gained attention and been gradually applied, with belief networks finding wide application in various fields. Deep belief networks are a type of deep learning network that uses Restricted Boltzmann Machines (RBMs) as its basic units. Compared to traditional backpropagation neural networks, RBMs are based on statistical mechanics. They possess stronger nonlinear generalization capabilities for unsupervised learning scenarios. It consists of visible units v and hidden units h, connected by weights w. Once the state (v, h) is determined, the energy of each RBM unit can be characterized as:
[0065]
[0066] In the formula: a and b are the biases of the visible layer and the hidden layer, respectively; θ is the optimal parameter. At this point, the joint probability distribution of (v, h) can be obtained as:
[0067]
[0068] In the formula: normalization factor z(θ)=∑ v,k e -E(v,h|θ) .
[0069] In practical applications, it is necessary to calculate the marginal distribution function of p(v,h|θ), that is:
[0070]
[0071] Based on its structural advantages, RBM can determine the activation probability of hidden units using the neuron states of the visible layer, thus:
[0072]
[0073] In the formula: σ is the process function. Similarly, based on the symmetry of its structure, the activation probability of the visible layer unit is obtained as:
[0074]
[0075] The parameter learning and updating of RBMs can be performed using the contrastive divergence (CD) algorithm. The parameter update method during the learning process is as follows:
[0076]
[0077] In the formula: ε is the update function.
[0078] Combining multiple RBM units with a single BP structure yields a deep belief network, which can learn deep features of data and significantly improve the network's nonlinear fitting ability.
[0079] The DBN network structure model includes hidden layers and an input layer. Their matrix operation relationships within the network are as follows:
[0080] h = wx(8
[0081] The matrices w and x can be represented in the following form:
[0082]
[0083] x = [x1 x2 … x] M ] T (10)
[0084] Because the computational structure of deep belief networks is relatively complex, the training process is prone to finding local optima and overfitting. Therefore, in designing the unmanned control strategy for bucket wheel excavators, this embodiment uses the Gaussian mutation-based firefly algorithm (LOGFA) to optimize the PID parameters. The specific process of improving the firefly algorithm optimization is as follows:
[0085] In the improved firefly algorithm, for a firefly population of size m in n-dimensional space, the position of the i-th population can be denoted as:
[0086]
[0087] In the improved firefly algorithm, four attributes were defined for this population: maximum fluorescence intensity I. 0i Relative brightness I i The mutual gravitational force F between objects ij (t) and the mass m of the object i The four attributes are calculated using the following formulas:
[0088]
[0089] In the formula: f(·) is the objective function used in the algorithm iteration process; γ is the light intensity absorption factor. Other parameters in the formula include r... ij G, f worst (x) i f hest (x) i The calculation formula is as follows:
[0090]
[0091] Based on the definition of the improved firefly algorithm above and the formula derivation above, the method for updating the firefly population location can be derived as follows:
[0092]
[0093] In the formula: a represents the iteration factor, and Δa is the decay coefficient during the iteration process.
[0094] This invention introduces an adjustable iteration factor, enabling the algorithm to adaptively change within the search range of feasible solutions during the optimization process, further improving iteration efficiency. This can overcome the drawback of the complex structure of deep belief networks.
[0095] In addition, to avoid stagnation in population evolution due to excessive iterations, this invention introduces a Gaussian mutation mechanism. This mechanism is configured such that if the population fails to evolve after five consecutive iterations, it mutates according to the strategy in Formula 15:
[0096] x′ i =x i +x i ×N(0,1) (15)
[0097] In the formula: N(0,1) is a Gaussian random variable.
[0098] The proposed automatic control strategy is applied when the bucket wheel excavator is in a fully automatic, unmanned state, repeating its work according to the set instruction 1. The system is designed to achieve management, scheduling, querying and other related functions of the bucket wheel excavator by writing the PID control algorithm into the PID control package in the PLC.
[0099] In the control system of a bucket wheel excavator, for the convenience of studying the problem, the actuator of the bucket wheel excavator can be simplified as an asynchronous motor as a simulation model. An asynchronous motor is a device that converts kinetic energy into mechanical energy, and the research on asynchronous motors is relatively mature. It can be studied using mathematical models. Under ideal conditions, the voltage vector control equation in the motor is:
[0100]
[0101] In the formula: s represents the stator in the motor; r represents the rotor; d and q are the d-axis and q-axis components of the magnetic flux linkage, respectively. When v rd =v rq When = 0, the magnetic flux linkage calculation formula in Formula 16 is as follows:
[0102]
[0103] After a simplified analysis of the asynchronous motor, the overall control strategy of the asynchronous motor closed-loop speed regulation system based on PID control can be obtained as follows: Figure 2 As shown in the figure, the solid box shows the control principle after writing PID in a traditional PLC, while the dashed box shows the optimization principle of the PID algorithm after using deep belief network and improved firefly algorithm.
[0104] The control strategy mentioned in this embodiment is an offline PLC control scheme. Its advantage is that it does not require real-time online data. The input and output of the PLC controller are extracted through the DBN network, and then the three parameters KP, KI, and KD in the PID are optimized based on the improved firefly algorithm. The PID controls the PLC through the continuous iteration of these three parameters.
[0105] Currently, a few thermal power plants in China still use local manual control methods. These systems rely on relay control systems driven by ordinary AC asynchronous motors. These systems have significant drawbacks: high cost, poor reliability, low automation, and poor operability. By incorporating programmable logic controllers (PLCs), frequency converters, soft starters, proximity switches, and photoelectric switches into the bucket wheel stacker-reclaimer control system, the number of relays and traditional control cables is significantly reduced. This results in a more rational system configuration, lower power consumption, and a substantial improvement in the stability of the bucket wheel stacker-reclaimer's operation. This enables semi-automatic operation of the bucket wheel stacker-reclaimer and increases production efficiency.
[0106] To meet the requirements of highly automated production in large power plants and modern ports, and to improve work efficiency, the operation of bucket wheel stacker-reclaimers must be automated and intelligent, with modernized communication and management technologies. Automatic control refers to the unmanned operation of the bucket wheel stacker-reclaimer, monitored by a monitoring system. This invention utilizes deep belief networks and an improved firefly algorithm to optimize three parameters in the PID programmable controller, enabling rapid convergence. Compared to the traditional fuzzy PID algorithm, the proposed algorithm significantly improves control accuracy; in this embodiment, the MAPE is 3.2576%, approximately 5% lower than the fuzzy PID control strategy, and the RMSE decreases to 0.0379.
[0107] Compared with traditional manual operation of bucket wheel excavators, the automatic control strategy proposed in this embodiment can greatly improve work efficiency, improve the working environment of the workforce and ensure the safety of the staff, enhance the level of automation of bucket wheel excavators in power plants, and further realize the modernization of power plant operations. Compared with existing automatic control schemes, this invention simplifies the algorithm structure and optimizes the iteration steps, which is of great significance for the research on optimization algorithms of programmable controllers.
[0108] In one embodiment, an automatic control system for a bucket wheel excavator based on a deep belief network is provided, comprising the following program modules:
[0109] Data Acquisition Module: Used to collect operating data of the bucket wheel excavator using a PLC controller, the operating data including the input and output parameters of the PLC controller; the collected operating data is used as the input of the DBN, and the network is trained through unsupervised learning so that it can identify the patterns and trends in the operation of the bucket wheel excavator;
[0110] Initialization module: Used to set initial parameters for the PID controller based on the initial operating conditions of the bucket wheel excavator;
[0111] The optimization module is used to define the attributes of the firefly population based on the firefly algorithm, calculate the position of each firefly through the objective function, and find the optimal solution of the PID parameters by simulating the firefly's bioluminescence behavior. In each iteration of the firefly population, the iteration factor is adjusted according to the current performance of the algorithm to adaptively adjust the search range. If the firefly population has not evolved in 5 consecutive iterations, Gaussian mutation is introduced to update the firefly positions according to a specific mutation strategy.
[0112] Update module: Used to update the parameters of the PID controller based on the optimized Firefly algorithm results, and write the updated PID parameters into the PLC to realize the automatic control of the bucket wheel excavator.
[0113] The specific implementation details of each module can be found in the above description of the limitations of the automatic control method for bucket wheel excavators based on deep belief networks, and will not be repeated here.
[0114] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 3 As shown, the computer device includes a processor, memory, and network interface connected via a system bus. The processor provides computing and control capabilities, and the network interface enables communication with external terminals via a network connection. The computer device loads and runs a computer program to implement the aforementioned automatic control method for bucket wheel excavators based on deep belief networks.
[0115] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.
[0116] In one embodiment, a computer program product is also provided, including a computer program / instructions that, when executed by a processor, relate to all or part of the processes in the methods of the above embodiments.
[0117] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
Claims
1. A bucket wheel machine automatic control method based on deep belief network, characterized in that: The method comprises: Using a PLC controller to collect bucket wheel machine operation data, the operation data includes input and output parameters of the PLC controller; using the collected operation data as input to a DBN, and training the network through unsupervised learning to enable it to identify patterns and trends in bucket wheel machine operation; According to the initial operation of the bucket wheel excavator, the initial parameters are set for the PID controller; The properties of the firefly population are defined based on the firefly algorithm, and the position of each firefly is calculated through the objective function. The optimal solution of the PID parameters is found by simulating the light-emitting behavior of the fireflies. In each iteration of the firefly population, the iteration factor is adjusted according to the current performance of the algorithm to adaptively adjust the search range. If the firefly population does not evolve in five consecutive iterations, Gaussian mutation is introduced to update the position of the fireflies according to a specific mutation strategy. According to the optimized firefly algorithm results, the parameters of the PID controller are updated, and the updated PID parameters are written into the PLC to realize automatic control of the bucket wheel excavator.
2. The method according to claim 1, characterized in that The PLC controller is used to collect the operation data of the bucket wheel machine, and the operation data includes the input and output parameters of the PLC controller including: Use the PLC system to collect bucket wheel machine operation data in real time, including but not limited to motor current, voltage, temperature, and vibration frequency; Providing accurate input data for DBN training ensures that the control strategy can make decisions based on actual operating conditions.
3. The method according to claim 1, characterized in that The PID controller sets the initial parameters specifically including: setting the proportional coefficient, the integral time and the differential time.
4. The method according to claim 1, characterized in that The updating of the parameters of the PID controller according to the optimized firefly algorithm results includes: The optimized PID parameters are written into the PLC to replace the original parameter settings. The control strategy of the bucket wheel excavator is adjusted according to the optimization results to improve the control accuracy and efficiency.
5. The method according to claim 1, characterized in that The updated PID parameters are written into the PLC to realize automatic control of the bucket wheel excavator, including: By using the updated PID parameters, the bucket wheel machine can be automatically controlled through PLC, and the intelligent control of the bucket wheel machine can be realized to improve production efficiency and safety.
6. The method according to claim 1, characterized in that After the updated PID parameters are written into the PLC to realize the automatic control of the bucket wheel excavator, the method further includes: Continuously monitor the operating status of the bucket wheel machine and feed back real-time data to the DBN network and firefly algorithm to achieve closed-loop control. Adjust the control strategy according to real-time data to ensure that the bucket wheel machine is always in the optimal operating state.
7. The method according to claim 1, characterized in that After writing the updated PID parameters into the PLC to realize the automatic control of the bucket wheel machine, the method also includes: regularly maintaining the system, checking the performance of the DBN network and the firefly algorithm, and adjusting and optimizing them as needed to ensure the long-term stable operation of the system.
8. A bucket wheel machine automatic control system based on deep belief network, characterized in that: include: Acquisition module: used to collect the operation data of the bucket wheel machine using the PLC controller, and the operation data includes the input and output parameters of the PLC controller; The collected operation data is used as the input of DBN, and the network is trained by unsupervised learning so that it can identify the patterns and trends in the operation of bucket wheel machine; Initialization module: used to set initial parameters for the PID controller according to the initial operation of the bucket wheel machine; Optimization module: used to define the properties of the firefly population based on the firefly algorithm, calculate the position of each firefly through the objective function, and find the optimal solution for PID parameters by simulating the luminous behavior of fireflies; In each iteration of the firefly population, the iteration factor is adjusted according to the current performance of the algorithm to adaptively adjust the search range; if the firefly population does not evolve in five consecutive iterations, Gaussian mutation is introduced to update the positions of the fireflies according to a specific mutation strategy; Update module: used to update the parameters of the PID controller according to the optimized firefly algorithm results, write the updated PID parameters into the PLC, and realize automatic control of the bucket wheel excavator.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.
10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to claims 1-7 are implemented.
