Biomass generator set control system based on self-adaptive adjustment

Abstract

The invention discloses a biomass generator set control system based on self-adaptive adjustment, belongs to the technical field of biomass power generation, and aims at solving the problems that in the prior art, under the condition that the external environment is frequently changed, a traditional control system cannot achieve real-time self-adaptive adjustment, so that unit operation is not stable, energy consumption is increased, and power consumption is low. And even a fault risk is caused. The technical effects are that according to the result provided by the state identification and prediction module, a multi-objective function taking the back pressure, the cooling water flow and the air suction rate as variables is constructed, an optimization boundary is set in combination with the operation constraint of the unit, and through communication with the PLC programmable logic controller, the air suction rate of the unit is improved. Linkage control over the rotating speed of a cooling water pump, the opening degree of an air exhaust valve, starting and stopping of a cooling fan and heat exchanger flow distribution key equipment is completed, continuous optimal management of unit operation parameters can be achieved through accurate, efficient and closed-loop execution control, and the safety and the intelligent level of overall operation are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of biomass power generation, and in particular to a biomass generator set control system based on adaptive regulation. Background Art

[0002] Biomass power generation, as a key form of renewable energy, has been widely adopted in industrial and regional energy supply. Biomass generators typically use high-moisture, organic fuels as their energy source. These generators suffer from poor combustion stability and frequent operating fluctuations, which can easily lead to abnormal turbine backpressure, fluctuating cooling efficiency, and low energy efficiency. Existing control systems generally rely on fixed parameters or manual adjustment methods, resulting in slow response and low adjustment accuracy, making them difficult to meet complex dynamic operational requirements.

[0003] When the external environment changes frequently, traditional control systems cannot achieve real-time adaptive adjustment, resulting in unstable unit operation, increased energy consumption, and even the risk of failure. To solve the above problems, we proposed a biomass generator set control system based on adaptive adjustment. Summary of the Invention

[0004] To this end, the present invention provides a biomass generator set control system based on adaptive regulation to solve the existing problem that when the external environment changes frequently, the traditional control system cannot achieve real-time adaptive regulation, resulting in unstable unit operation, increased energy consumption, and even the risk of failure.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] According to a first aspect of the present invention, a biomass power generation set control system based on adaptive regulation includes the following contents:

[0007] Data acquisition module, used to collect various data of the generator set;

[0008] The state recognition and prediction module uses a multi-source fusion model to identify the current operating state and whether there is low vacuum, overheating or abnormal load conditions. It also uses the LSTM time series prediction model to predict the trend of key parameter changes in the short term in the future, supporting forward-looking control.

[0009] The fault-tolerant judgment module introduces a confidence mechanism and anomaly detection algorithm to identify data drift, model anomalies, and sensor failures;

[0010] The adaptive and control module, based on a multi-objective optimization algorithm, combines the three goals of power generation efficiency, cooling energy consumption, and operational safety margin to automatically generate the optimal regulation strategy, including cooling water flow regulation, load distribution, and extraction ratio adjustment. It converts the regulation strategy into specific controllable instructions and jointly executes various actuators, including cooling water pumps, extraction valves, fans, and heat exchangers, to achieve precise control of unit operating parameters.

[0011] The monitoring and display module displays the unit's operating status, alarm information, forecast trends and adjustment effects in real time through a graphical interface, and provides a manual intervention interface.

[0012] Furthermore, the data acquisition module collects the operating data of the biomass power generation unit, deploys a variety of sensors, and collects key parameters such as turbine back pressure, extraction volume, cooling water flow and water temperature in real time. Environmental data sensors are deployed near the operation of the biomass power generation unit to collect environmental temperature, atmospheric pressure, humidity, and inlet water quality data.

[0013] Furthermore, the state recognition and prediction module uses a support vector machine machine learning classification algorithm to build a training model by offline labeling of historical operation data, and performs feature mapping and classification judgment on the current collected data states. It can maintain high recognition accuracy under conditions of high data dimension and frequent signal fluctuations, predict future changes in key operating parameters in advance, judge potential operation risks and optimize the adjustment time window.

[0014] Furthermore, the fault-tolerant judgment module judges various abnormal situations in the model output or operation process based on the data processed by the state recognition and prediction module, ensures the stability, reliability and robustness of the system operation, sets the confidence interval, and automatically marks the current result as "low confidence" when the confidence level is lower than the threshold, enters the fault-tolerant processing logic, and cross-validates the key parameters.

[0015] Furthermore, the adaptive control module integrates multi-objective optimization algorithms and real-time operating status perception to achieve dynamic adjustment and intelligent optimization of parameters of data processed by the state recognition and prediction modules, solves strategies through genetic algorithms, and quickly searches for optimal or near-optimal adjustment schemes. The optimization results are then mapped to specific execution instructions, and the optimal adjustment strategy generated by adaptive adjustment is converted into specific control instructions. By communicating with the PLC programmable logic controller, the linkage control of key equipment such as cooling water pump speed, exhaust valve opening, cooling fan start and stop, and heat exchanger flow distribution is completed.

[0016] Furthermore, the monitoring and display module is connected to the data streams of each sensor, the state recognition results and the control execution feedback in real time, and a visual display interface is set up, which can be displayed in real time in the form of charts, curves and heat maps.

[0017] The present invention has the following advantages:

[0018] By deploying a variety of sensors, key parameters such as turbine back pressure, extraction volume, cooling water flow, and water temperature are collected in real time. Turbine back pressure is a core indicator for measuring condenser vacuum and unit expansion end efficiency. Using the support vector machine machine learning classification algorithm, a training model is constructed by offline annotating historical operating data. Feature mapping and classification estimation are performed on the current input state data to predict future changes in key operating parameters in advance, identify potential operating risks, and optimize the adjustment time window. By calculating the confidence level of the model identification or prediction output results, the probability output distribution, the size of the prediction residual, and the output fluctuation range, a confidence interval is set. Once a "data logic conflict" or "abnormal deviation" occurs, an abnormal flag is triggered. If the output results fluctuate significantly, frequently reverse, or non-physical jumps are found, the model may be judged to be overfitting or abnormal, and the backup model mode is entered or the adjustment instruction output is terminated.

[0019] Based on the results provided by the state identification and prediction module, a multi-objective function with back pressure, cooling water flow, and extraction volume as variables is constructed, and the optimization boundary is set in combination with the unit's operating constraints. By communicating with the PLC programmable logic controller, the system completes the linkage control of key equipment such as cooling water pump speed, extraction valve opening, cooling fan start and stop, and heat exchanger flow distribution. Through precise, efficient, closed-loop execution control, the system can achieve continuous optimization management of the unit's operating parameters, improving the safety and intelligence level of the overall operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] To more intuitively illustrate the prior art and the present application, exemplary drawings are provided below. It should be understood that the specific shapes and structures shown in the drawings should not generally be considered as limiting conditions for implementing the present application. For example, those skilled in the art, based on the technical concepts disclosed in the present application and the exemplary drawings, are capable of easily making routine adjustments or further optimizations to the addition / reduction / attribution division, specific shapes, positional relationships, connection methods, and dimensional ratios of certain units (components).

[0021] Figure 1 A module diagram of a biomass power generation set control system based on adaptive regulation provided in some embodiments of the present invention. DETAILED DESCRIPTION

[0022] The following describes the implementation of the present invention using specific embodiments. Those skilled in the art will readily understand the other advantages and benefits of the present invention from the disclosure herein. Obviously, the embodiments described are only a portion of the present invention, not all of it. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are intended to fall within the scope of protection of the present invention.

[0023] Example 1

[0024] Figure 1 As shown, a biomass power generation set control system based on adaptive regulation in an embodiment of the first aspect of the present invention includes the following contents:

[0025] Data acquisition module, used to collect various data of the generator set;

[0026] Biomass power generation unit operation data collection, through the deployment of a variety of sensors, real-time collection of key parameters such as turbine back pressure, extraction volume, cooling water flow and water temperature. Turbine back pressure is the core indicator for measuring condenser vacuum and unit expansion end efficiency. Increased back pressure usually indicates the occurrence of low vacuum state, which directly affects power generation efficiency and system stability. Extraction volume reflects the energy distribution of different extraction stages of the turbine. Its changes have a significant impact on exhaust conditions and back pressure, and are an important basis for evaluating thermal system regulation strategies. Cooling water flow determines the heat exchange capacity of the condenser and is a key variable affecting vacuum maintenance and cooling tower load regulation. Water temperature, especially the inlet and outlet water temperatures of the cooling water, can be used to judge the cooling effect and the operating efficiency of the cooling system.

[0027] For data collection of operating environment, environmental data sensors are deployed near the operation of biomass power generation units. Ambient temperature, atmospheric pressure, humidity, and inlet water quality are also crucial. Ambient temperature will directly affect the cooling water temperature and the heat exchange efficiency of the condenser, and is an important external factor for judging the vacuum change trend; atmospheric pressure fluctuations will affect the reference benchmark of the condenser absolute pressure, and thus affect the back pressure judgment and vacuum adjustment strategy; ambient humidity reflects the evaporation capacity and cooling efficiency changes of the cooling tower to a certain extent; and the water quality parameters of the cooling water source (salt content, turbidity) are related to the scaling tendency and heat transfer performance of the heat exchanger, which in turn affects the system's vacuum maintenance ability.

[0028] The technical effect achieved by the above scheme is: by deploying multiple sensors, the key parameters of turbine back pressure, suction volume, cooling water flow and water temperature are collected in real time. Turbine back pressure is the core indicator for measuring the vacuum degree of the condenser and the efficiency of the expansion end of the unit.

[0029] The state recognition and prediction module uses a multi-source fusion model to identify the current operating state and whether there is low vacuum, overheating or abnormal load conditions. It also uses the LSTM time series prediction model to predict the trend of key parameter changes in the short term in the future, supporting forward-looking control.

[0030] State recognition is used to make real-time judgments on the operating conditions of the biomass power generation unit at the current moment of operation, identifying whether there are typical operating problems such as low vacuum, high back pressure, insufficient cooling, and abnormal load. Using the support vector machine machine learning classification algorithm, a training model is constructed by offline annotation of historical operating data. Feature mapping and classification judgments are performed on the currently collected data states. The algorithm has excellent nonlinear fitting capabilities and noise resistance, and can maintain high recognition accuracy under conditions of high data dimensions and frequent signal fluctuations.

[0031] State prediction: Based on the study of historical operating trends and time series modeling, it predicts future changes in key operating parameters in advance, identifies potential operating risks, and optimizes adjustment time windows. This module focuses on addressing the timing issues of back pressure increase trends, water temperature fluctuations, and system load change trends, improving the system's forward-looking control capabilities.

[0032] The long-short-term memory neural network can effectively capture long-term dependencies and is suitable for processing multivariable, nonlinear, and trend-sensitive operating data in power generation systems. It has strong predictive stability and generalization capabilities. The prediction results can not only be used to determine whether low vacuum or high load conditions will occur in the future, but also provide a basis for parameter adjustment in advance for the multi-objective adjustment module, realizing "pre-adjustment" logic.

[0033] Through the coordinated operation of the identification and prediction modules, the system can achieve accurate perception of the current operating status and effective prediction of future development trends, thereby improving the intelligence, stability and response speed of the entire control system.

[0034] Fault-tolerant judgment module: Introduces a confidence mechanism and anomaly detection algorithm to identify data drift, model anomalies, and sensor failures to ensure system robustness;

[0035] Fault-tolerant judgment: Based on the data processed by the state recognition and prediction module, the system judges various abnormal situations in the model output or operating state, including sensor failure, model drift, and sudden changes in operating conditions. This prevents the system from erroneously triggering control instructions or policy adjustments due to erroneous information, ensuring the stability, reliability, and robustness of the system operation. In combination with the confidence assessment mechanism, the credibility of the recognition results is calculated and secondary verification is performed by comparing with historical similar operating conditions to avoid misjudgments caused by short-term disturbances.

[0036] By calculating the confidence level of the model identification or prediction output results, the probability output distribution, the size of the prediction residual, the output fluctuation range, and setting the confidence interval, when the confidence level is lower than the threshold, the current result is automatically marked as "low confidence", and the fault-tolerant processing logic is switched to cross-validation of key parameters. By setting the physical law logic to determine whether the data are consistent, once a "data logic conflict" or "abnormal deviation" occurs, the abnormal mark is triggered, and the stability analysis of multiple consecutive rounds of identification or prediction results is performed. If it is found that the output results fluctuate greatly, frequently reverse, or non-physical jumps occur, it is judged that the model may be overfitting or abnormal, and the backup model mode is entered or the adjustment instruction output is terminated.

[0037] The technical effects achieved by the above scheme are as follows: using the support vector machine machine learning classification algorithm, building a training model by offline labeling of historical operation data, performing feature mapping and classification judgment on the current input state data, predicting future changes in key operating parameters in advance, judging potential operating risks and optimizing the adjustment time window, and calculating the confidence of the model identification or prediction output results, the probability output distribution, the size of the prediction residual, the output fluctuation amplitude, and setting the confidence interval. Once a "data logic conflict" or "abnormal deviation" occurs, the abnormal mark is triggered. If the output result is found to fluctuate greatly, frequently reverse or non-physical jumps occur, it is judged that the model may be overfitting or abnormal, and the backup model mode is entered or the adjustment instruction output is terminated.

[0038] Adaptive and control module: Based on a multi-objective optimization algorithm, it combines the triple goals of power generation efficiency, cooling energy consumption, and operational safety margin to automatically generate the optimal regulation strategy, including cooling water flow regulation, load distribution, and extraction ratio adjustment. It converts the regulation strategy into specific controllable instructions and jointly executes various actuators, including cooling water pumps, extraction valves, fans, and heat exchangers, to achieve precise control of unit operating parameters.

[0039] Adaptive regulation integrates multi-objective optimization algorithms with real-time operating status perception to achieve dynamic parameter adjustment and intelligent optimization of data processed by the state recognition and prediction module. Maximizing power generation efficiency, minimizing cooling energy consumption, and ensuring system operational safety are the core optimization goals. First, based on the results provided by the state recognition and prediction module, a multi-objective function is constructed with back pressure, cooling water flow, and extraction volume as variables. The optimization boundaries are set in conjunction with the unit's operating constraints, and a genetic algorithm is used to solve the strategy, rapidly searching for the optimal or near-optimal regulation solution. The optimization results are then mapped into specific execution instructions.

[0040] Control execution converts the optimal adjustment strategy generated by adaptive adjustment into specific control instructions. By communicating with the PLC programmable logic controller, it completes the linkage control of key equipment such as cooling water pump speed, exhaust valve opening, cooling fan start and stop, and heat exchanger flow distribution. Through precise, efficient, closed-loop execution control, the system can achieve continuous optimization management of unit operating parameters, improving the safety and intelligence level of overall operation.

[0041] The monitoring and display module displays the unit's operating status, alarm information, forecast trends, and adjustment effects in real time through a graphical interface, and provides a manual intervention interface to facilitate operator decision-making assistance and operation supervision;

[0042] Equipment monitoring: Through real-time access to sensor data streams, status recognition results, and control execution feedback, dynamic monitoring of the full-process operation status of the biomass power generation unit is achieved;

[0043] Data display, setting up a visual display interface, can display key operating parameters and vacuum state change trends in real time in the form of charts, curves, and heat maps. By integrating the output results and recommended strategies of the optimization algorithm, the interface can show users multiple adjustment suggestions in the form of decision prompts, thereby improving the efficiency of human-computer collaboration; the display interface can also support authority classification and operation record tracking functions to enhance the standardization and traceability of operation management.

[0044] The technical effect achieved by the above scheme is as follows: based on the results provided by the state identification and prediction module, a multi-objective function with back pressure, cooling water flow and exhaust volume as variables is constructed, and the optimization boundary is set in combination with the operating constraints of the unit. By communicating with the PLC programmable logic controller, the linkage control of key equipment such as cooling water pump speed, exhaust valve opening, cooling fan start and stop, and heat exchanger flow distribution is completed. Through precise, efficient, and closed-loop execution control, the system can achieve continuous optimization management of the unit's operating parameters, improve the overall safety and intelligence level of operation, and the data display makes it easy for operators to quickly understand the current system status; the interface can simultaneously display historical data comparison, model identification results and prediction and warning information, which helps operation and maintenance personnel to promptly detect potential low vacuum risks and make intervention decisions.

[0045] The technical features of the above embodiments can be combined arbitrarily (as long as there is no contradiction in the combination of these technical features). In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described; these embodiments that are not explicitly written should also be considered to be within the scope of this specification.

Claims

1. A biomass generator set control system based on adaptive regulation, characterized in that: Includes the following: Data acquisition module, used to collect various data of the generator set; The state recognition and prediction module uses a multi-source fusion model to identify the current operating state and whether there is low vacuum, overheating or abnormal load conditions. It also uses the LSTM time series prediction model to predict the trend of key parameter changes in the short term in the future, supporting forward-looking control. The fault-tolerant judgment module introduces a confidence mechanism and anomaly detection algorithm to identify data drift, model anomalies, and sensor failures; The adaptive and control module, based on a multi-objective optimization algorithm, combines the three goals of power generation efficiency, cooling energy consumption, and operational safety margin to automatically generate the optimal regulation strategy, including cooling water flow regulation, load distribution, and extraction ratio adjustment. It converts the regulation strategy into specific controllable instructions and jointly executes various actuators, including cooling water pumps, extraction valves, fans, and heat exchangers, to achieve precise control of unit operating parameters. The monitoring and display module displays the unit's operating status, alarm information, forecast trends and adjustment effects in real time through a graphical interface, and provides a manual intervention interface.

2. The biomass power generation set control system based on adaptive regulation according to claim 1 is characterized in that: The data acquisition module collects the operating data of the biomass generator set, deploys multiple sensors, and collects key parameters such as turbine back pressure, extraction volume, cooling water flow and water temperature in real time. Environmental data sensors are deployed near the operation of the biomass generator set to collect ambient temperature, atmospheric pressure, humidity, and inlet water quality data.

3. The biomass power generation set control system based on adaptive regulation according to claim 1, characterized in that: The state recognition and prediction module uses a support vector machine machine learning classification algorithm to build a training model by offline labeling of historical operation data, and performs feature mapping and classification judgment on the current collected data states. It can maintain high recognition accuracy under conditions of high data dimension and frequent signal fluctuations, predict future changes in key operating parameters in advance, judge potential operation risks and optimize the adjustment time window.

4. The biomass power generation set control system based on adaptive regulation according to claim 1, characterized in that: The fault-tolerant judgment module judges various abnormal situations in the model output or operation process based on the data processed by the state recognition and prediction module to ensure the stability, reliability and robustness of the system operation. It sets a confidence interval and automatically marks the current result as "low confidence" when the confidence level falls below the threshold. The fault-tolerant processing logic is switched to cross-validation of key parameters.

5. The biomass power generation set control system based on adaptive regulation according to claim 1, characterized in that: The adaptive control module integrates multi-objective optimization algorithms with real-time operating status perception to achieve dynamic adjustment and intelligent optimization of parameters for data processed by the state recognition and prediction modules. It uses genetic algorithms to solve strategies and quickly search for optimal or near-optimal adjustment schemes. The optimization results are then mapped into specific execution instructions. The optimal adjustment strategy generated by adaptive adjustment is converted into specific control instructions. By communicating with the PLC programmable logic controller, the linkage control of key equipment such as the cooling water pump speed, the exhaust valve opening, the cooling fan start and stop, and the heat exchanger flow distribution is completed.

6. The biomass power generation set control system based on adaptive regulation according to claim 1, characterized in that: The monitoring and display module is connected to the sensor data stream, state recognition results and control execution feedback in real time, and sets a visual display interface, which can be displayed in real time in the form of charts, curves and heat maps.

Images