Method for reducing abrasion of pneumatic ash conveying pipeline of power plant and optimizing energy consumption based on real-time monitoring

Abstract

The invention relates to the field of household kitchens, and discloses a power plant pneumatic ash conveying pipeline abrasion reduction and energy consumption optimization method based on real-time monitoring, which comprises the following steps: S1, arranging a sensor group at a node of an ash conveying pipeline; s2, preprocessing the sensor data through a data processing system; s3, constructing a wear prediction model based on the preprocessed data; s4, based on the pipeline abrasion trend, the airflow speed and pressure are dynamically adjusted; s5, adjusting the equipment power according to the influence of the speed and pressure of the adjusted airflow on the sensor data; s6, feeding back the data of each step to a control system; and S7, establishing a pipeline three-dimensional model. Through real-time collection of pipeline operation state data, data processing and a wear prediction model, the airflow speed and pressure are dynamically adjusted, and the equipment power is adaptively adjusted according to the change of the airflow, so that the wear of the pipeline can be obviously slowed down while the conveying efficiency is kept, and the service life of the pipeline is prolonged.

Description

Technical Field

[0001] The invention relates to the field of household kitchens, and in particular to a method for alleviating wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring. Background Art

[0002] The pneumatic ash conveying pipeline of a power plant is a pipeline system used to transport and convey solid waste such as dust and fly ash in the boiler of a power plant. Its main function is to transport the solid waste in the boiler to a designated storage or treatment device through airflow during the coal-fired power generation process. This pneumatic ash conveying system is designed to reduce dust emissions generated by coal combustion, prevent it from polluting the environment, and improve the efficiency of dust treatment.

[0003] During the pneumatic ash conveying process, the pipeline is prone to wear due to the long-term impact and friction of high-speed airflow, especially in bends and joints. In addition, the pressure fluctuations caused by high-speed airflow will also increase the energy consumption of the pipeline, resulting in a decrease in the overall efficiency of the system. If effective wear control and energy consumption management measures are not taken, premature wear of the pipeline will increase maintenance costs and downtime risks, and energy consumption problems will lead to energy waste.

[0004] While existing technologies can monitor the wear of pneumatic ash conveying pipelines to a certain extent, they usually lack real-time data processing and prediction capabilities, making it difficult to identify high-wear risk areas in a timely manner, and fail to effectively analyze and warn of wear trends. In addition, existing technologies rely on fixed modes for airflow control and energy consumption management, making it difficult to dynamically adjust airflow speed and pressure according to real-time operating conditions. As a result, the system's energy efficiency level is low, and the pipeline wear control effect is poor. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides a method for reducing wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring, which solves the problems in the prior art of lack of real-time data processing and prediction capabilities, difficulty in timely identifying high wear risk areas, and difficulty in dynamically adjusting airflow velocity and pressure according to real-time operating conditions.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A method for reducing wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring, comprising the following steps:

[0007] S1. Arrange a sensor group at the node of the ash conveying pipeline to collect pipeline operation status data in real time;

[0008] S2, preprocessing the sensor data through the data processing system, including denoising, time synchronization and data storage;

[0009] S3. Build a wear prediction model based on the preprocessed data and predict the pipeline wear trend through deep learning algorithm;

[0010] S4, dynamically adjust air flow speed and pressure based on pipeline wear trend;

[0011] S5, adjusting the power of the device according to the influence of the speed and pressure of the adjusted airflow on the sensor data;

[0012] S6, feeding back the data of each step to the control system to form an adaptive closed-loop control for real-time evaluation of system performance;

[0013] S7. A three-dimensional pipeline model is established based on the sensor group data and the equipment information to display the pipeline status in real time.

[0014] Preferably, the sensor group in S1 includes a flow rate sensor, a pressure sensor, a temperature sensor and a vibration sensor, and the sensor group is arranged at the bends, interfaces, horizontal sections and vertical sections of the pipeline, for real-time detection of the flow rate, pressure, temperature and vibration changes in the bends, interfaces, horizontal sections and vertical sections of the ash conveying pipeline.

[0015] Preferably, in S2, the data processing uses a median filter method to perform denoising on the collected data to eliminate the interference of environmental noise, and the denoising formula is:

[0016] X′ t =median(X t-1 ,X t ,X t+1 )

[0017] Among them, X t represents the sensor data collected at time t, X′ t is the processed data, median(X t-1 ,X t ,X t+1 ) is the median of the data at time t-1, t, and t+1, which is the denoised data value.

[0018] Preferably, the wear prediction model in S3 adopts a long short-term memory network to predict the wear trend of the pipeline by inputting the data feature vector of the sensor, wherein the feature vector includes air flow velocity, pressure change rate, temperature fluctuation and vibration root mean square value:

[0019]

[0020] Among them, F t is the sensor data feature vector, is the average velocity of the airflow, Δp is the pressure change rate per unit time, ΔT is the temperature fluctuation, that is, the difference between the maximum temperature and the minimum temperature, and RMS(a) is the root mean square value of the vibration;

[0021] The average velocity of the airflow is calculated by the following formula:

[0022]

[0023] Where: v i is the flow rate reading of the i-th sensor at time t, and N is the total number of flow rate sensors;

[0024] The pressure change rate is calculated by the following formula:

[0025]

[0026] Where: P t and P t+1 are the pressure values ​​at time t and t+1 respectively, and ΔT is the time interval;

[0027] The temperature fluctuation is calculated by the following formula:

[0028] ΔT=T max -T min

[0029] Where: T max is the maximum temperature, T min is the minimum temperature;

[0030] The root mean square value of the vibration is calculated by the following formula:

[0031]

[0032] Among them: a i is the vibration data of the i-th vibration sensor, and N is the total number of vibration sensors.

[0033] Preferably, the mean square error is used as the loss function to optimize the prediction results. The loss function is:

[0034]

[0035] Among them, L(θ) is the loss function value, θ is the model parameter, N is the total number of data samples, T is the number of time steps, is the wear value at the tth moment predicted by the model, y t is the actual wear value at moment t.

[0036] Preferably, in S4, airflow control is implemented by a deep network model, and the deep network updates the airflow velocity and pressure according to the airflow state to balance wear reduction and delivery efficiency. The update formula is:

[0037]

[0038] Among them, Q(S,A) represents the Q value of selecting action A under the current state S. The Q value reflects the total reward expected after taking this action under the current state. The higher the Q value, the better the effect of selecting this action. S is the current airflow state, including the current flow rate, pressure, temperature and vibration parameters. A is the current action, that is, the specific adjustment of the airflow speed or pressure. α is the learning rate, which is used to control the step size of the value update. The larger the learning rate, the faster the system updates the new experience; conversely, the learning speed becomes slower. R is the immediate reward value, which is the reward obtained by selecting action A under the current state S. The reward setting is based on the impact of airflow regulation on pipeline wear. Positive rewards are given when pipeline wear is reduced or transportation efficiency is improved, and negative rewards are given when wear is increased or transportation efficiency is reduced. γ is a discount factor, which is used to weigh the impact of future rewards on current decisions. The value is between 0 and 1. The closer the discount factor is to 1, the more the system pays attention to long-term effects. If the discount factor is smaller, it means that the system pays more attention to immediate effects. It represents the maximum Q value of all possible actions in the next state S′. This value represents the optimal action that the system can take in the next state and its corresponding Q value, which is used to evaluate future rewards in the current state.

[0039] Preferably, the objective function of the airflow optimization is:

[0040]

[0041] in, is the optimization of V and P, W is the wear impact measure of the target optimization, V is the airflow velocity, P is the airflow pressure, α and β are weight coefficients used to balance the impact of velocity and pressure on wear.

[0042] Preferably, in S5, the power of the equipment is adjusted for energy consumption optimization management, including adaptive frequency conversion control of the fan and the motor, and specifically dynamically adjusting the speed of the fan and the motor through the following formula:

[0043]

[0044] Among them, f new and m new are the adjusted fan and motor powers, respectively; f and m are the adjusted fan speed and motor power, η is the adjustment coefficient, and is the partial derivative of energy consumption E with respect to f and m.

[0045] Preferably, in S6, the feedback system realizes real-time evaluation of system performance by calculating the wear reduction rate and the energy consumption optimization rate, and the calculation formula is as follows:

[0046]

[0047] Among them, R w is the wear reduction rate, R e is the energy consumption optimization rate, ΔW is the improvement value of wear optimization, defined as the initial wear amount W initial The difference between the initial energy consumption and the optimized wear amount, ΔE, is the improvement value of energy consumption optimization, which is defined as the initial energy consumption E initial The difference between the optimized energy consumption and the

[0048] Preferably, in S7, a three-dimensional model of the pipeline is established through a data analysis and visualization module, and the model displays real-time data of the sensor, pipeline wear conditions and equipment operating status in three-dimensional form, and is updated in real time as the data changes, so as to display the dynamic operating status.

[0049] The present invention provides a method for reducing wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring. It has the following beneficial effects:

[0050] 1. The present invention arranges a sensor group at the ash conveying pipeline node to realize real-time collection of pipeline operation status data, dynamically adjusts the airflow speed and pressure through data processing and wear prediction model, and adaptively adjusts the equipment power according to the change of airflow. Therefore, it can significantly reduce the wear of the pipeline while maintaining the transportation efficiency, thereby extending the service life of the pipeline.

[0051] 2. The present invention dynamically adjusts the fan and motor power through adaptive frequency conversion control, and gradually optimizes the system energy consumption by adjusting the step size according to the impact of energy consumption on the fan speed and motor power. It can not only reduce the energy waste caused by excessive air flow speed or pressure, but also adaptively adjust the power according to actual working conditions, effectively achieving energy saving and consumption reduction.

[0052] 3. The present invention establishes a three-dimensional pipeline model through a data analysis and visualization module to display sensor data, pipeline wear and equipment operating status in real time. The three-dimensional model is updated in real time as the data changes, providing operation and maintenance personnel with a dynamic and intuitive operating status display, which helps to promptly discover and deal with pipeline wear and equipment problems, thereby improving the system's operating reliability and maintenance efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] Embodiment 1:

[0056] Please refer to the attached Figure 1 The embodiment of the present invention provides a method for reducing wear and optimizing energy consumption of a pneumatic ash conveying pipeline in a power plant based on real-time monitoring, comprising the following steps:

[0057] S1. Arrange a sensor group at the node of the ash conveying pipeline to collect pipeline operation status data in real time;

[0058] S2, preprocessing the sensor data through the data processing system, including denoising, time synchronization and data storage;

[0059] S3. Build a wear prediction model based on the preprocessed data and predict the pipeline wear trend through deep learning algorithm;

[0060] S4, dynamically adjust air flow speed and pressure based on pipeline wear trend;

[0061] S5, adjusting the power of the device according to the influence of the speed and pressure of the adjusted airflow on the sensor data;

[0062] S6, feeding back the data of each step to the control system to form an adaptive closed-loop control for real-time evaluation of system performance;

[0063] S7. A three-dimensional pipeline model is established based on the sensor group data and the equipment information to display the pipeline status in real time.

[0064] The sensor group in S1 includes a flow rate sensor, a pressure sensor, a temperature sensor and a vibration sensor, and the sensor group is arranged at the bends, interfaces, horizontal sections and vertical sections of the pipeline to detect the flow rate, pressure, temperature and vibration changes of the bends, interfaces, horizontal sections and vertical sections of the ash conveying pipeline in real time.

[0065] In the present invention, first, a sensor group is arranged at the key nodes of the ash conveying pipeline to collect pipeline operation status data in real time. The sensor group includes flow rate, pressure, temperature and vibration sensors, which can comprehensively monitor the changes in operating parameters inside the pipeline, especially in key locations such as bends and interfaces, and can more accurately capture wear and stress changes. Through the cooperation of these sensors, the data processing system performs denoising, time synchronization and data storage on the real-time collected data, providing a high-quality data foundation for subsequent wear prediction and energy consumption optimization. Subsequently, the system constructs a wear prediction model based on the cleaned data, and uses a deep learning algorithm to analyze the wear trend of the pipeline, so that the system can respond in time before the wear problem worsens.

[0066] And based on the prediction of pipeline wear trends, the system dynamically adjusts the airflow speed and pressure to achieve more precise control. Specifically, based on the feedback of the adjusted airflow speed and pressure on the sensor data, the system automatically adjusts the equipment power to form an adaptive closed-loop control mechanism to maintain optimal energy efficiency. At the same time, this data is fed back to the control system to evaluate the system performance in real time. In order to enable operation and maintenance personnel to understand the pipeline status in real time, the system establishes a three-dimensional model of the pipeline based on sensor data and equipment operation information. The model can dynamically display the pipeline status, including real-time monitoring data, wear trends, and equipment adjustment results. Therefore, operation and maintenance personnel can not only intuitively view the real-time operating status of the pipeline, but also adjust strategies based on the model to improve the convenience and accuracy of system maintenance.

[0067] In S2, the data processing uses the median filter method to denoise the collected data to eliminate the interference of environmental noise. The denoising formula is:

[0068] X′ t =median(X t-1 ,X t ,X t+1 )

[0069] Among them, X t represents the sensor data collected at time t, X′ t is the processed data, median(X t-1 ,X t ,X t+1 ) is the median of the data at time t-1, t, and t+1, which is the denoised data value.

[0070] In the present invention, the collected sensor data is denoised by a data processing module, and a median filtering method is used to eliminate the interference of environmental noise on the sensor data. During the operation of the ash conveying pipeline, environmental factors such as vibration, wind flow fluctuations, etc. may cause noise interference to the sensor data, resulting in abnormalities or deviations in the collected data, which will affect the accuracy of the wear prediction model. Therefore, the purpose of denoising is to ensure the stability and reliability of the sensor data. The median filtering method compares multiple data points within a period of time, takes the median as the representative value at that moment, and effectively filters out short-term mutations or extreme values, thereby obtaining a smoother data curve. In this way, the accuracy of the data can be greatly improved, misjudgment caused by noise can be avoided, and more accurate data support can be provided for subsequent data analysis, prediction and regulation.

[0071] In addition, the data processing system not only plays a key role in denoising, but also can realize synchronous processing of data. Since multiple sensors in the pipeline are distributed at different nodes, there will be certain time differences in the data collected by these nodes. Through the synchronization mechanism of the data processing system, the acquisition time of each sensor can be aligned to ensure that all sensor data reflects the status of the pipeline in the same time period. The synchronized data can more accurately reflect the real-time operating status of the pipeline and avoid data deviation problems caused by time differences. The accuracy of the data is further improved, laying a high-quality foundation for subsequent wear prediction and energy consumption optimization, ensuring that the entire pipeline system operates under efficient and stable monitoring.

[0072] The wear prediction model in S3 uses a long short-term memory network to predict the wear trend of the pipeline by inputting the data feature vector of the sensor. The feature vector includes air flow velocity, pressure change rate, temperature fluctuation and vibration root mean square value:

[0073]

[0074] Among them, F t is the sensor data feature vector, is the average velocity of the airflow, Δp is the pressure change rate per unit time, ΔT is the temperature fluctuation, that is, the difference between the maximum temperature and the minimum temperature, and RMS(a) is the root mean square value of the vibration;

[0075] The average velocity of the air flow is calculated by the following formula:

[0076]

[0077] Where: v i is the flow rate reading of the i-th sensor at time t, and N is the total number of flow rate sensors;

[0078] The pressure change rate is calculated using the following formula:

[0079]

[0080] Where: P t and P t+1 are the pressure values ​​at time t and t+1 respectively, and ΔT is the time interval;

[0081] The temperature fluctuation is calculated by the following formula:

[0082] ΔT=T max -T min

[0083] Where: T max is the maximum temperature, T min is the minimum temperature;

[0084] The root mean square value of vibration is calculated by the following formula:

[0085]

[0086] Among them: a i is the vibration data of the i-th vibration sensor, and N is the total number of vibration sensors.

[0087] In the present invention, the wear prediction model uses a long short-term memory network (LSTM) to analyze the wear trend of the ash conveying pipeline. The LSTM model has the advantage of processing time series data and can capture the time dependency in the data. It is suitable for predicting complex problems such as pipeline wear that are affected by multiple factors such as airflow velocity, pressure, and temperature. In terms of model input, the LSTM model uses the feature vector of sensor data as the main input, and predicts the future pipeline wear trend by analyzing the temporal variation pattern of each feature. The feature vector contains multi-dimensional data such as airflow velocity, pressure change rate, temperature fluctuation, and vibration root mean square value. By comprehensively considering these parameters, the model can more comprehensively understand and quantify the operating status of the pipeline, providing reliable data support for wear prediction.

[0088] Specifically, the air flow velocity can reflect the speed of gas flow in the ash conveying system. Excessive flow velocity may lead to aggravated wear, while appropriate flow velocity helps to control the degree of wear. The pressure change rate is used to describe the fluctuation of pressure. By monitoring the pressure change per unit time, the system can identify the wear tendency caused by external load or system resistance change. Temperature fluctuation reflects the thermal environment of the gas and the inner wall of the pipeline. When the temperature difference is large, the thermal expansion and contraction of the pipeline material may generate stress, thereby accelerating wear. In addition, the root mean square value of vibration reflects the mechanical vibration state of the pipeline. This feature can help identify the vibration intensity caused by air flow turbulence or pipeline friction, which is particularly important for analyzing the wear at bends and interfaces. Through these characteristic parameters, the LSTM model can more accurately understand the complex dynamic behavior of the pipeline system and achieve efficient prediction of wear trends. The wear prediction model performs deep learning under multi-parameter input, and generates a trend prediction for future wear by identifying the change pattern of each feature over time. This prediction can identify high-wear risk areas in advance, thereby providing timely warnings for operation and maintenance personnel, helping them take necessary maintenance measures before the wear reaches a critical state. This will enable forward-looking prediction methods, which will not only improve the stability of the system, but also effectively extend the service life of the equipment, and achieve more refined energy efficiency management and equipment maintenance.

[0089] The mean square error is used as the loss function to optimize the prediction results. The loss function is:

[0090]

[0091] Among them, L(θ) is the loss function value, θ is the model parameter, N is the total number of data samples, T is the number of time steps, is the wear value at the tth moment predicted by the model, y t is the actual wear value at moment t.

[0092] In S4, airflow control is implemented through a deep network model. The deep network updates the airflow velocity and pressure according to the airflow state to balance wear reduction and transportation efficiency. The update formula is:

[0093]

[0094] Among them, Q(S,A) represents the Q value of selecting action A under the current state S. The Q value reflects the total reward expected after taking this action under the current state. The higher the Q value, the better the effect of selecting this action. S is the current airflow state, including the current flow rate, pressure, temperature and vibration parameters. A is the current action, that is, the specific adjustment of the airflow speed or pressure. α is the learning rate, which is used to control the step size of the value update. The larger the learning rate, the faster the system updates the new experience; conversely, the learning speed becomes slower. R is the immediate reward value, which is the reward obtained by selecting action A under the current state S. The reward setting is based on the impact of airflow regulation on pipeline wear. Positive rewards are given when pipeline wear is reduced or transportation efficiency is improved, and negative rewards are given when wear is increased or transportation efficiency is reduced. γ is a discount factor, which is used to weigh the impact of future rewards on current decisions. The value is between 0 and 1. The closer the discount factor is to 1, the more the system pays attention to long-term effects. If the discount factor is smaller, it means that the system pays more attention to immediate effects. It represents the maximum Q value of all possible actions in the next state S′. This value represents the optimal action that the system can take in the next state and its corresponding Q value, which is used to evaluate future rewards in the current state.

[0095] The objective function of airflow optimization is:

[0096]

[0097] in, is the optimization of V and P, W is the wear impact measure of the target optimization, V is the airflow velocity, P is the airflow pressure, α and β are weight coefficients used to balance the impact of velocity and pressure on wear.

[0098] In the wear prediction model of the invention, the mean square error (MSE) is used as the loss function to optimize the prediction accuracy of the model. The mean square error measures the prediction error of the model at each time step by calculating the square difference between the predicted value and the actual value. By minimizing the loss function value, the model can gradually learn and optimize the parameters so that the prediction results are closer to the true value. Therefore, it can effectively capture significant errors in the prediction deviation, especially the high penalty for large deviations, which helps the model to minimize large prediction errors and improve prediction accuracy. During the optimization process, as the number of training iterations increases, the loss function value of the model gradually decreases, indicating that the learning effect of the model has been gradually improved, and finally a relatively stable wear trend prediction result is achieved. Based on this optimization method, the system can effectively identify high wear risks of pipelines, which helps to prevent possible failures in advance.

[0099] In terms of airflow control, the present invention uses a deep network model to dynamically adjust the airflow speed and pressure, aiming to balance the reduction of pipeline wear and transportation efficiency. The deep network gradually learns the optimal control strategy by inputting airflow state parameters, including flow rate, pressure, temperature, vibration and other factors. The system selects operations in each state and evaluates the effect of the current operation based on the immediate reward value. If the operation helps to reduce pipeline wear or improve transportation efficiency, a positive reward is given; on the contrary, if it aggravates wear or affects efficiency, a negative reward is given. In addition, the system balances immediate benefits and long-term benefits through discount factors, which enables the system to consider long-term transportation efficiency and energy efficiency while reducing current wear. With the continuous iteration of the deep network model, the system gradually forms the optimal airflow control strategy, which not only effectively controls pipeline wear, but also ensures the stability and economy of the overall operation of the system, and achieves the target effect of airflow optimization.

[0100] In S5, the equipment power is adjusted for energy consumption optimization management, including adaptive frequency conversion control of fans and motors. Specifically, the fan and motor speeds are dynamically adjusted through the following formula:

[0101]

[0102] Among them, f new and m new are the adjusted fan and motor powers, respectively; f and m are the adjusted fan speed and motor power, η is the adjustment coefficient, and is the partial derivative of energy consumption E with respect to f and m.

[0103] In the present invention, energy consumption is optimized and managed by dynamically adjusting the power of the equipment, especially for the adaptive frequency conversion control of the fan and the motor. The fan and the motor play an important role in promoting and supporting the ash conveying system, but during operation, the load of the fan and the motor will also fluctuate due to changes in the ash conveying demand. To adapt to this change, the system dynamically adjusts the speed of the fan and the motor according to the data feedback of real-time monitoring, so that the equipment can match the current working condition requirements, thereby achieving optimal energy efficiency. The control strategy continuously updates the power of the fan and the motor through an adaptive frequency conversion control algorithm to reduce unnecessary energy consumption. For example, when the conveying demand decreases, the system can automatically reduce the speed of the fan and the motor to reduce energy consumption; when the demand increases, the system can increase the speed of the fan and the motor to ensure the conveying efficiency. This adjustment process combines the factors affecting the power of the fan and the motor with the energy consumption, and controls the adjustment step size through a sensitive adjustment coefficient to ensure that the power change process can both respond to system requirements and maintain stability. This intelligent dynamic adjustment method not only improves the energy efficiency of the system, but also effectively reduces energy waste in long-term operation, extends the service life of the equipment, further reduces the operation and maintenance costs of the system, and achieves high efficiency and energy saving effects.

[0104] In S6, the feedback system realizes real-time evaluation of system performance by calculating the wear reduction rate and energy consumption optimization rate. The calculation formula is as follows:

[0105]

[0106] Among them, R w is the wear reduction rate, R e is the energy consumption optimization rate, ΔW is the improvement value of wear optimization, defined as the initial wear amount W initial The difference between the initial energy consumption and the optimized wear amount, ΔE, is the improvement value of energy consumption optimization, which is defined as the initial energy consumption E initial The difference between the optimized energy consumption and the

[0107] In the present invention, the feedback system calculates the wear reduction rate and the energy consumption optimization rate to achieve real-time evaluation of the performance of the entire system. The feedback mechanism can continuously monitor and quantify the effectiveness of wear control and energy consumption management during the operation of the pipeline, and comprehensively evaluate the optimization effect of the system in a data-driven manner. Specifically, the wear reduction rate intuitively shows the degree of improvement of the wear trend by comparing the difference between the optimized wear amount and the initial wear amount; it can help operation and maintenance personnel judge the actual effect of pipeline wear management, and facilitate adjustment of strategies when wear reduction is not ideal. The energy consumption optimization rate quantifies the actual effectiveness of the equipment in improving energy efficiency by comparing the difference between the initial energy consumption and the optimized energy consumption. This indicator directly reflects the impact of the adaptive control of the fan and the motor on energy consumption, so that the system can accurately track the improvement of energy efficiency. If the energy consumption optimization rate shows a significant improvement, it indicates that the system has an ideal energy-saving effect in the current operating mode; otherwise, it may be necessary to further adjust the power management strategy, which not only improves the visualization of the system operation, so that the operation and maintenance personnel can clearly understand the performance of the current system, but also supports quick decision-making, ensuring that the system always operates in an efficient and low-consumption manner, thereby achieving the best operation and maintenance effect.

[0108] In S7, the three-dimensional model of the pipeline is established through the data analysis and visualization module. The model displays the real-time data of the sensor, pipeline wear and equipment operating status in three dimensions, and is updated in real time as the data changes, so as to display the dynamic operating status.

[0109] In the present invention, a three-dimensional model of the pipeline is established through a data analysis and visualization module, and three-dimensional and dynamic monitoring of the pipeline system is realized. Since it is built based on real-time sensor data, the structure and operation status of the pipeline can be intuitively displayed in the virtual space, and the real-time changes of parameters such as flow rate, pressure, temperature and vibration can be synchronously reflected. With the continuous updating of data, the three-dimensional model will also be dynamically adjusted to display the wear conditions inside and outside the pipeline and the operation status of the equipment in real time. Through this visual method, the operation and maintenance personnel can view the overall operation of the pipeline system at a glance, and quickly identify the areas with severe wear or equipment with abnormal operation. The establishment of the three-dimensional model not only improves the readability of the data, but also enables the operation and maintenance personnel to intuitively locate potential problem areas, so as to facilitate timely measures. At the same time, it helps to track the wear development trend of the pipeline in the long term, assist managers to make scientific decisions based on historical data and real-time status when formulating maintenance plans, and further improve the operation reliability and efficiency of the pipeline system. By providing this comprehensive and intuitive dynamic monitoring view, the three-dimensional model provides strong support for intelligent operation and maintenance management, and improves the maintenance convenience and response speed of the entire system.

[0110] Embodiment 2:

[0111] In this embodiment, after completing data collection, denoising, synchronization and storage, the system can call previously stored historical data for real-time comparison when necessary. By retrieving and analyzing previous data, the system extracts features in the current data that are similar to historical situations to predict potential risks that may occur under current conditions. For example, if the feature vector of the current data matches the historical high wear state in terms of air flow velocity, pressure, temperature fluctuation or vibration mode, the system will automatically identify this dangerous trend and generate a warning message. The warning message will directly mark the high-risk area on the three-dimensional model of the pipeline so that operation and maintenance personnel can quickly identify and locate possible dangerous points.

[0112] In addition, the 3D model not only has the function of real-time display, but also can enter the "preview mode" under system requirements to simulate the operating conditions in the future. In the preview mode, the operation and maintenance personnel can manually adjust the power parameters of the motor and fan to observe the impact of these control measures on the operation of the pipeline. The 3D model will simultaneously display the operating status of the pipeline under different power configurations, including air flow velocity, pressure distribution, temperature changes and vibration conditions, so as to help the operation and maintenance personnel evaluate the effects of different adjustment strategies. At the same time, the system's learning model will record and analyze the data feedback during the preview process, summarize the simulation results under different power settings, and identify the optimal adjustment strategy. These strategies will be gradually optimized and applied to actual operation. The learning model will continue to learn and improve under actual working conditions, thereby continuously enhancing its ability to adjust the power of motors and fans, and ultimately improving the stability and energy efficiency of the system, achieving the dual goals of wear reduction and energy consumption optimization.

[0113] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for reducing wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring, characterized in that: The following steps are involved: S1. Arrange a sensor group at the node of the ash conveying pipeline to collect pipeline operation status data in real time; S2, preprocessing the sensor data through the data processing system, including denoising, time synchronization and data storage; S3. Build a wear prediction model based on the preprocessed data and predict the pipeline wear trend through deep learning algorithm; S4, dynamically adjust air flow speed and pressure based on pipeline wear trend; S5, adjusting the power of the device according to the influence of the speed and pressure of the adjusted airflow on the sensor data; S6, feeding back the data of each step to the control system to form an adaptive closed-loop control for real-time evaluation of system performance; S7. A three-dimensional pipeline model is established based on the sensor group data and the equipment information to display the pipeline status in real time.

2. The method for reducing wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring according to claim 1 is characterized in that: The sensor group in S1 includes a flow rate sensor, a pressure sensor, a temperature sensor and a vibration sensor, and the sensor group is arranged at the bends, interfaces, horizontal sections and vertical sections of the pipeline, and is used to detect the flow rate, pressure, temperature and vibration changes of the bends, interfaces, horizontal sections and vertical sections of the ash conveying pipeline in real time.

3. The method for reducing wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring according to claim 1 is characterized in that: In S2, the data processing uses the median filtering method to perform denoising on the collected data to eliminate the interference of environmental noise. The denoising formula is: X′ t =median(X t-1 ,X t ,X t+1 ) Among them, X t represents the sensor data collected at time t, X′ t is the processed data, median(X t-1 , X t , X t+1 ) is the median of the data at time t-1, t, and t+1, which is the denoised data value.

4. The method for reducing wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring according to claim 1 is characterized in that: The wear prediction model in S3 uses a long short-term memory network to predict the wear trend of the pipeline by inputting the data feature vector of the sensor. The feature vector includes air flow velocity, pressure change rate, temperature fluctuation and vibration root mean square value: Among them, F t is the sensor data feature vector, is the average velocity of the airflow, Δp is the pressure change rate per unit time, ΔT is the temperature fluctuation, that is, the difference between the maximum temperature and the minimum temperature, and RMS(a) is the root mean square value of the vibration; The average velocity of the airflow is calculated by the following formula: Where: v i is the flow rate reading of the i-th sensor at time t, and N is the total number of flow rate sensors; The pressure change rate is calculated by the following formula: Where: P t and P t+1 are the pressure values ​​at time t and t+1 respectively, and Δt is the time interval; The temperature fluctuation is calculated by the following formula: ΔT=T max -T min Where: T max is the maximum temperature, T min is the minimum temperature; The root mean square value of the vibration is calculated by the following formula: Among them: a i is the vibration data of the i-th vibration sensor, and N is the total number of vibration sensors.

5. The method for reducing wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring according to claim 4 is characterized in that: The mean square error is used as the loss function to optimize the prediction results. The loss function is: Among them, L(θ) is the loss function value, θ is the model parameter, N is the total number of data samples, T is the number of time steps, is the wear value at the tth moment predicted by the model, y t is the actual wear value at moment t.

6. The method for reducing wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring according to claim 1 is characterized in that: In S4, airflow control is implemented through a deep network model. The deep network updates the airflow velocity and pressure according to the airflow state to balance wear reduction and delivery efficiency. The update formula is: Among them, Q(S, A) represents the Q value of selecting action A under the current state S. The Q value reflects the total reward expected to be obtained after taking this action under the current state. The higher the Q value, the better the effect of selecting this action. S is the current airflow state, including the current flow rate, pressure, temperature and vibration parameters. A is the current action, that is, the specific adjustment of the airflow speed or pressure. α is the learning rate, which is used to control the step size of the value update. The larger the learning rate, the faster the system updates the new experience; conversely, the learning speed becomes slower. R is the immediate reward value, which is the reward obtained by selecting action A under the current state S. The reward setting is based on the impact of airflow regulation on pipeline wear. Positive rewards are given when pipeline wear is reduced or transportation efficiency is improved, and negative rewards are given when wear is increased or transportation efficiency is reduced. γ is a discount factor, which is used to weigh the impact of future rewards on current decisions. The value is between 0 and 1. The closer the discount factor is to 1, the more the system pays attention to long-term effects. If the discount factor is smaller, it means that the system pays more attention to immediate effects. It represents the maximum Q value of all possible actions in the next state S′. This value represents the optimal action that the system can take in the next state and its corresponding Q value, which is used to evaluate future rewards in the current state.

7. The method for reducing wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring according to claim 6 is characterized in that: The objective function of the airflow optimization is: in, is the optimization of V and P, W is the wear impact measure of the target optimization, V is the airflow velocity, P is the airflow pressure, α and β are weight coefficients used to balance the impact of velocity and pressure on wear.

8. The method for reducing wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring according to claim 1 is characterized in that: In S5, the power of the equipment is adjusted for energy consumption optimization management, including adaptive frequency conversion control of the fan and the motor, and specifically dynamically adjusting the fan and motor speeds through the following formula: Among them, f new and m new are the adjusted fan and motor powers, respectively; f and m are the adjusted fan speed and motor power, η is the adjustment coefficient, and is the partial derivative of energy consumption E with respect to f and m.

9. The method for reducing wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring according to claim 1, characterized in that: In S6, the feedback system realizes real-time evaluation of system performance by calculating the wear reduction rate and energy consumption optimization rate. The calculation formula is as follows: Among them, R w is the wear reduction rate, R e is the energy consumption optimization rate, ΔW is the improvement value of wear optimization, defined as the initial wear amount W initial The difference between the initial energy consumption and the optimized wear amount, ΔE, is the improvement value of energy consumption optimization, which is defined as the initial energy consumption E initial The difference between the optimized energy consumption and the 10. The method for reducing wear and optimizing energy consumption of pneumatic ash conveying pipelines in power plants based on real-time monitoring according to claim 1, characterized in that: In S7, a three-dimensional model of the pipeline is established through a data analysis and visualization module. The model displays real-time data of sensors, pipeline wear conditions, and equipment operating status in three dimensions, and is updated in real time as the data changes, so as to display the dynamic operating status.

Images