Heat recovery system and method for lubricating oil system of power plant turboset

Abstract

The invention discloses a power plant turboset lubricating oil system heat recovery system and method, and relates to the technical field of turbine power generation equipment.The method comprises the steps that real-time operation data and historical operation logs of multiple systems of a target scene are collected in real time; performing dynamic performance baseline identification on a heat exchanger between the lubricating oil system and the heat transfer medium recovery system to obtain a real-time performance baseline representing the current heat transfer performance; predicting a thermal state change trend of the lubricating oil system in a future preset time period; calculating a feedforward control quantity for adjusting the flow of the heat transfer medium; and generating a composite control instruction of the heat transfer medium flow, and executing heat recovery control. The technical problems that an existing power plant turboset lubricating oil system is low in heat recovery efficiency, cooling equipment is high in energy consumption, and lubricating oil temperature control lags behind are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of steam turbine power generation equipment, in particular to a power plant steam turbine unit lubricating oil system heat recovery system and method. BACKGROUND

[0002] With the development of various power generation technologies in power plants towards high parameters and large capacities, the heat management problem of the steam turbine unit lubricating oil system is increasingly prominent. The traditional lubricating oil system usually adopts water cooling or air cooling to directly dissipate heat, resulting in waste of heat, and the energy consumption of the cooling equipment continues to rise, not only restricting the improvement of heat recovery efficiency, but also increasing the operation and maintenance cost of the unit. Even if a simple PID automatic control technology is adopted to recover heat, there are still technical problems of temperature control lag of the lubricating oil. SUMMARY

[0003] The embodiments of the present application provide a power plant steam turbine unit lubricating oil system heat recovery system and method, which solve the technical problems of low heat recovery efficiency, high energy consumption of cooling equipment and temperature control lag of the lubricating oil of the existing lubricating oil system.

[0004] The technical solutions of the present application to solve the above technical problems are as follows: In a first aspect, the present application provides a power plant steam turbine unit lubricating oil system heat recovery system, comprising: A data acquisition unit is configured to acquire real-time operation data and historical operation logs of multiple systems in a target scene in real time. A performance baseline updating unit is configured to perform dynamic performance baseline identification on a heat exchanger between a lubricating oil system and a heat medium recovery system based on the historical operation logs, and acquire a real-time performance baseline representing current heat transfer performance. A thermal dynamic prediction unit is configured to predict a thermal state change trend of the lubricating oil system in a future preset period based on the real-time operation data. A feedforward control calculation unit is configured to calculate a feedforward control amount for adjusting the flow of the heat medium in combination with the real-time performance baseline and the thermal state change trend, with the goal of maintaining stable temperature before the lubricating oil tank. A compound control output unit is configured to generate a compound control instruction of the flow of the heat medium according to the feedforward control amount, and execute heat recovery control.

[0005] In a second aspect, the present application provides a power plant steam turbine unit lubricating oil system heat recovery method, comprising: Acquiring real-time operation data and historical operation logs of multiple systems in a target scene in real time. Performing dynamic performance baseline identification on a heat exchanger between a lubricating oil system and a heat medium recovery system based on the historical operation logs, and acquiring a real-time performance baseline representing current heat transfer performance. predict a thermal state change trend of the lubricating oil system in a future preset period based on the real-time operation data; combine the real-time performance baseline and the thermal state change trend to calculate a feedforward control amount for adjusting the heat transfer medium flow rate, aiming to maintain the temperature stability before the lubricating oil tank; generate a composite control instruction of the heat transfer medium flow rate according to the feedforward control amount, and perform heat recovery control.

[0006] The application provides one or more technical solutions, at least having the following technical effects or advantages: The application provides a power plant steam turbine unit lubricating oil system heat recovery system and method. First, the data acquisition unit acquires real-time operation data and historical operation logs of the lubricating oil system, the heat transfer medium recovery system and the main system in the target scene in real time. Second, the performance baseline updating unit performs dynamic performance baseline identification on the heat exchanger based on the historical operation logs to obtain a real-time performance baseline representing the current heat transfer performance, reflecting the actual heat transfer capacity of the heat exchanger under different working conditions. Third, the thermal dynamic prediction unit predicts the thermal state change trend of the lubricating oil system in a future preset period using the real-time operation data, so that the system can perceive the possible fluctuations of the lubricating oil temperature in advance and change from passive control to active prevention. Fourth, the feedforward control calculation unit combines the real-time performance baseline and the thermal state change trend to calculate a feedforward control amount aiming to maintain the temperature stability before the lubricating oil tank, and pre-adjusts the heat transfer medium flow rate from the source to reduce the possibility of temperature fluctuations. Finally, the composite control output unit generates a composite control instruction according to the feedforward control amount and performs heat recovery control, which not only considers the predicted thermal state change but also combines the actual temperature deviation in the operation for feedback adjustment, thereby improving the heat recovery efficiency.

[0007] Through the above technical solutions, the application reduces the energy consumption of the cooling equipment and effectively improves the problem of temperature control hysteresis of the lubricating oil, achieving the dual goals of efficient recovery of the lubricating oil system heat and temperature control. BRIEF DESCRIPTION OF DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0009] Figure 1 is a structural schematic diagram of a power plant steam turbine unit lubricating oil system heat recovery system provided by the embodiments of the application; Figure 2This is a schematic flowchart of a heat recovery method for a power plant turbine unit lubrication oil system provided in an embodiment of this application.

[0010] The components represented by each number in the attached diagram are explained below: Data acquisition unit 11, performance baseline update unit 12, thermal dynamic prediction unit 13, feedforward control calculation unit 14, composite control output unit 15. Detailed Implementation

[0011] This application provides a heat recovery system and method for the lubricating oil system of a power plant turbine unit, which addresses the technical problems of low heat recovery efficiency, high energy consumption of cooling equipment, and lag in lubricating oil temperature control in existing lubricating oil systems.

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0013] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0014] In the description of this application, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use this application. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid unnecessarily obscuring the description of this application. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0015] Example 1, as Figure 1 As shown in the figure, this application provides a heat recovery system for the lubricating oil system of a power plant turbine unit, including: The data acquisition unit 11 is configured to acquire real-time operation data and historical operation logs of the target scene multi-system in real time. In the embodiment of the application, the temperature sensor and the volume flow meter arranged at the outlet of the main engine of the lubricating oil system and in front of the lubricating oil tank are used to collect the lubricating oil temperature and the lubricating oil volume flow.

[0016] Meanwhile, the temperature sensor and the volume flow meter arranged before and after the heat transfer medium recovery system are used to collect the pre-recovery temperature, the post-recovery temperature and the volume flow of the heat transfer medium.

[0017] In addition, the DCS system or other data storage servers of the power plant are communicated to acquire the near operation logs and the historical operation logs containing the historical records of the above parameters, and the main engine load parameters of the main engine system.

[0018] Specifically, the data acquisition unit 11 in the system comprises: acquire the lubricating oil operation parameters and the historical lubricating oil operation logs of the lubricating oil system; acquire the medium operation parameters and the historical medium operation logs of the heat transfer medium recovery system; acquire the main engine load parameters and the historical main engine operation logs of the main engine system; output the lubricating oil operation parameters, the medium operation parameters and the main engine load parameters as the real-time operation data; output the historical lubricating oil operation logs, the historical medium operation logs and the historical main engine operation logs as the historical operation logs.

[0019] In the embodiment of the application, first, the data connection with the sensor network of the lubricating oil system is established to acquire the lubricating oil operation parameters in real time, including the temperature of the lubricating oil at different key nodes such as the outlet of the main engine bearing, the inlet and outlet of the cooler and the inlet of the lubricating oil tank, the real-time volume flow of the lubricating oil, the pressure of the lubricating oil and the liquid level of the lubricating oil tank.

[0020] Meanwhile, the historical records related to the lubricating oil are retrieved and integrated from the historical database to form the historical lubricating oil operation logs, including the change of the above lubricating oil operation parameters in different time periods, the system start-stop records, the maintenance records and the abnormal working condition information that has occurred.

[0021] Secondly, for the heat transfer medium recovery system, the temperature sensor, the volume flow meter and the pressure sensor arranged on the pipeline of the system are used to collect the inlet temperature of the heat transfer medium before entering the heat exchanger, the outlet temperature of the heat transfer medium after flowing out of the heat exchanger, the real-time volume flow of the medium in the pipeline and the operation pressure of the medium.

[0022] Correspondingly, the historical medium operation log containing the historical data of the above-mentioned medium operation parameters, the medium supplement record, the historical operation state of the related pump group, and the like is obtained from a historical storage module of the heat transfer medium recovery system itself.

[0023] Again, data interaction is performed with the control center or DCS system of the power plant main engine system to obtain real-time main engine load parameters, such as the power generation amount of the main engine, the steam flow, the rotating speed or shaft power of the steam turbine, and the like, which can represent the current operation load level of the main engine.

[0024] Meanwhile, the variation curve of the main engine load parameter in different historical periods, the main engine start-stop record, the load adjustment instruction record, the operation state history of the key auxiliary machine of the main engine, and the like are extracted from the historical database of the main engine system to form a historical main engine operation log.

[0025] Finally, the real-time collected lubricating oil operation parameters, medium operation parameters, and main engine load parameters are subjected to data format standardization processing and timestamp alignment to ensure consistency and accuracy in the time dimension, and then are integrated and packaged to output real-time operation data required for subsequent processing.

[0026] Specifically, for the historical lubricating oil operation log, the historical medium operation log, and the historical main engine operation log, data cleaning is performed to remove invalid data and outliers, reasonable interpolation or marking is performed on missing data, and data aggregation is performed according to a preset time granularity, such as every hour or every day, and finally the processed three types of historical logs are integrated and output as the historical operation log required by the system.

[0027] The performance baseline updating unit 12 is configured to perform dynamic performance baseline identification on the heat exchanger between the lubricating oil system and the heat transfer medium recovery system based on the historical operation log to obtain a real-time performance baseline representing the current heat transfer performance. The lubricating oil system and the heat transfer medium recovery system exchange heat through the heat exchange component arranged in front of the lubricating oil tank.

[0028] In the embodiment, first, based on the integrated historical operation log, sample data of the heat exchanger under normal working conditions is screened out. An improved least squares support vector machine algorithm is used to train the screened sample data to construct a mapping model between the heat transfer performance of the heat exchanger and various influencing factors.

[0029] Then, based on the real-time collected lubricating oil operation parameters and medium operation parameters, the logarithmic mean temperature difference and the actual heat exchange amount under the current working condition are calculated and input into the trained mapping model to obtain the predicted value of the current heat transfer performance.

[0030] Meanwhile, combined with the benchmark heat transfer performance data under the same working condition in the historical operation log, the deviation correction coefficient of the predicted value is dynamically adjusted through the sliding window mean comparison method, and a real-time performance baseline reflecting the current actual heat transfer capacity of the heat exchanger is generated. The real-time performance baseline is continuously updated as the running time progresses and the working condition changes.

[0031] The heat exchanger assembly adopts a plate heat exchanger, and the plate sheet material is selected from 316L stainless steel, which has corrosion resistance and heat conduction performance, and can adapt to the working condition of long-term contact of lubricating oil and heat transfer medium. The plate sheet of the plate heat exchanger is designed in a herringbone shape, which can enhance the turbulence degree of the fluid and improve the heat transfer coefficient. Compared with the traditional shell-and-tube heat exchanger, the heat transfer efficiency is improved by 30%-40% under the same heat exchange area.

[0032] Meanwhile, the sealing gasket of the heat exchanger is made of ethylene-propylene-diene rubber material, which has a temperature resistance range of -40℃ to 150℃, ensuring good sealing performance and service life within the normal working temperature range of the lubricating oil system. In addition, the heat exchanger is provided with a detachable structure, which facilitates regular cleaning of oil stains and impurities on the surface of the plate, maintains stable heat transfer performance, and when the gasket or plate needs to be replaced, the equipment does not need to be disassembled as a whole, greatly shortening the maintenance time.

[0033] Specifically, the performance baseline updating unit 12 in the system comprises: According to the preset neighborhood time window, the adjacent data is extracted from the historical operation log to obtain the adjacent operation log; According to the adjacent operation log, the first lubricating oil temperature at the host end outlet, the second lubricating oil temperature before the lubricating oil tank and the lubricating oil volume flow of the lubricating oil system are obtained, and the first real-time parameter set is output; According to the adjacent operation log, the pre-recovery temperature, the post-recovery temperature and the medium volume flow of the heat transfer medium recovery system are obtained, and the second real-time parameter set is output; According to the first real-time parameter set and the second real-time parameter set, the inverse operation is performed combined with the preset heat transfer basic model to obtain the real-time performance baseline.

[0034] In the embodiment of the application, first, a preset neighborhood time window is set, for example, the past 24 hours or the last 1000 groups of sampling data, the adjacent operation log closest to the current working condition is extracted from the historical operation log, and the representative historical data segment is selected to ensure the accuracy and timeliness of the subsequent performance baseline identification.

[0035] Secondly, the first lubricating oil temperature at the host end outlet, the second lubricating oil temperature before the lubricating oil tank and the lubricating oil volume flow of the lubricating oil system are selected from the extracted adjacent operation log.

[0036] The first oil temperature is the initial temperature of the lubricating oil after leaving the main bearing, and the second oil temperature is the temperature of the lubricating oil before entering the oil tank. The first oil temperature, the second oil temperature and the oil volume flow rate are combined and output as a first real-time parameter set, which reflects the initial state, target state and flow condition of the lubricating oil in the heat exchange process.

[0037] At the same time, the pre-recovery temperature, the post-recovery temperature and the medium volume flow rate of the heat transfer medium recovery system are also obtained from the adjacent operation log.

[0038] The pre-recovery temperature is the temperature of the heat transfer medium before entering the heat exchanger, and the post-recovery temperature is the temperature of the heat transfer medium after absorbing the heat of the lubricating oil and leaving the heat exchanger. The pre-recovery temperature, the post-recovery temperature and the medium volume flow rate are combined and output as a second real-time parameter set, which reflects the state change and flow condition of the heat transfer medium in the heat exchange process.

[0039] Finally, the first real-time parameter set and the second real-time parameter set are substituted into a preset heat transfer basic model, such as a heat exchanger heat transfer calculation model established based on the ε-NTU method or the logarithmic mean temperature difference method, to solve the real-time heat transfer coefficient or total heat transfer thermal resistance and other key performance parameters of the heat exchanger through inversion operation.

[0040] Further, based on the historical operation log, a dynamic performance baseline of the heat exchanger between the lubricating oil system and the heat transfer medium recovery system is identified to obtain a real-time performance baseline representing the current heat transfer performance, including: obtaining a medium-side operation parameter sequence of a current time window; inputting the medium-side operation parameter sequence into a pre-constructed heat transfer performance attenuation mapping model to obtain a heat transfer performance attenuation factor; calculating the real-time performance baseline based on the performance attenuation factor and the baseline heat transfer performance calculated based on design data.

[0041] In the embodiments of the present application, first, the medium-side operation parameter sequence of the heat transfer medium recovery system is extracted from the historical operation log corresponding to the current time window. The sequence includes continuous data points of the heat transfer medium volume flow rate, the pre-recovery temperature, the post-recovery temperature and the medium operating pressure collected at a fixed time interval, for example, every 1 minute, within a preset time window, for example, the past 1 hour.

[0042] Secondly, the extracted medium-side operation parameter sequence is input into a pre-constructed heat transfer performance attenuation mapping model. The model is trained based on the performance attenuation law of the heat exchanger at different operation periods in the historical data, for example, by analyzing the correlation between the medium flow rate change, the temperature difference change and the heat transfer coefficient attenuation, and using neural network or multivariate regression method to construct.

[0043] Further, after receiving the medium-side operation parameter sequence, the model calculates the change trend, fluctuation range and deviation degree from historical baseline data of each index in the parameter sequence, and outputs a heat exchange performance attenuation factor capable of quantifying the attenuation degree of the current heat exchange performance of the heat exchanger. The value range of the factor is usually 0 to 1, and the closer to 1, the smaller the performance attenuation and the closer to the initial design state of the heat transfer effect.

[0044] Finally, according to the baseline heat exchange performance parameters of the heat exchanger under the rated operating condition provided in the design data, such as the design heat transfer coefficient and the design heat exchange capacity, the baseline heat exchange performance x (1-heat exchange performance attenuation factor) is calculated to obtain a real-time performance baseline considering the current performance attenuation. The baseline can accurately reflect the actual heat transfer capacity of the heat exchanger under the current operating condition.

[0045] The attenuation mapping model is constructed, including: Obtaining historical medium operation logs and extracting the medium-side operation parameter sequence, wherein the medium-side operation parameter sequence includes the average flow rate and the average inlet temperature of the heat transfer medium; Obtaining the historical average heat transfer coefficient having time sequence consistency with the medium-side operation parameter sequence from the historical operation logs, wherein the historical average heat transfer coefficient is obtained by inverse calculation based on a preset heat transfer basic model; Taking the feature vector composed of the average flow rate and the average inlet temperature of the heat transfer medium as the model input and taking the corresponding historical average heat transfer coefficient as the model output target, a preset regression model is supervised and trained to obtain the heat transfer performance attenuation mapping model.

[0046] In the embodiments of the present application, first, sample data containing a complete operating cycle are selected from historical medium operation logs, for example, medium operation parameter records of 24 hours per day in the past 6 months are selected to ensure the comprehensiveness and representativeness of the data. For each preset time sub-window, such as every hour, the average value of the volume flow rate of the heat transfer medium in the time period is calculated as the average flow rate, and the average value of the inlet temperature of the heat transfer medium before entering the heat exchanger is calculated as the average inlet temperature. Two average values are combined to form a feature vector in the medium-side operation parameter sequence.

[0047] Secondly, for the historical operation data completely corresponding to the above feature vector in time, according to the preset heat transfer basic model, the actual heat transfer coefficient of the heat exchanger in the time period is inversely calculated by using the oil running parameters in the time period, including the first oil temperature, the second oil temperature, the oil volume flow rate and the medium operation parameters, including the temperature before recovery, the temperature after recovery and the medium volume flow rate.

[0048] Subsequently, a support vector regression model is selected, and a training set and a validation set are divided in a ratio of 7:3, the model is trained in a supervised learning manner, all extracted feature vectors are taken as input sample sets, i.e., average flow and average inlet temperature are taken as inputs, the number of nodes of an input layer is equal to the dimension of input features, for example, if the input layer has 2 features, the input layer has 2 nodes; the corresponding historical average heat transfer coefficient is taken as an output sample set, an output layer generally does not use an activation function, for example, if the output has 2 nodes, a continuous value is directly output; 1-3 layers of hidden layers are set, the number of nodes of each layer is adjusted through experiments, for example, 64, 32, etc., and a ReLU is selected as an activation function.

[0049] In the training process, the hyperparameters of the model are continuously adjusted to minimize the prediction error on the validation set, such as the root mean square error, when the validation set loss does not decrease for 5 consecutive rounds, the training process is automatically terminated, and a heat transfer performance attenuation mapping model capable of mapping the relationship between the medium side operating parameters and the heat transfer coefficient is obtained.

[0050] The thermal dynamic prediction unit 13 is configured to predict a thermal state change trend of the lubricating oil system in a future preset period based on the real-time operation data. Further, the thermal state change trend of the lubricating oil system in the future preset period is predicted based on the real-time operation data, and before that, the method comprises: The lubricating oil temperature time series data and the host system host load parameter time series data of the lubricating oil system are extracted based on the historical operation log. The lubricating oil temperature trend and the host load trend are obtained by respectively performing trend analysis on the lubricating oil temperature time series data and the host load parameter time series data. The trend alignment result is used to determine a lubricating oil-load typical delay. The data reorganization based on time series adjustment is performed on the lubricating oil temperature time series data and the host load parameter time series data according to the lubricating oil-load typical delay, and a delay removal mapping relationship is established. The load-lubricating oil temperature mapping model based on regression analysis is established based on the data reorganization result.

[0051] In the embodiments of the present application, first, the lubricating oil temperature time series data of the lubricating oil system in a plurality of continuous complete operation cycles is extracted from the historical operation log, for example, the time series of temperature parameters such as the lubricating oil tank inlet temperature and the host outlet oil temperature collected every 10 seconds, and the host load parameter time series data of the host system in the same time period, such as the time series data of the steam turbine shaft power or the generator output power recorded every 5 seconds.

[0052] Further, the extracted lubricating oil temperature time series data is preprocessed, including removing high-frequency noise using a moving average method, filling occasional short-time data gaps through linear interpolation, and identifying and removing abnormal jump values using a 3σ criterion to ensure the stationarity and continuity of the data sequence. Similarly, the same preprocessing procedure is performed on the host load parameter time series data to eliminate data abnormalities caused by sensor fluctuations or communication interference, obtaining the purified lubricating oil temperature time series and host load time series.

[0053] Secondly, trend analysis is performed on the preprocessed lubricating oil temperature time series data and host load parameter time series data respectively. A sliding window trend extraction algorithm is used, with a window size of 1 hour. The linear fitting slope of the lubricating oil temperature in each window is calculated as the temperature trend indicator. A positive slope indicates an upward trend, and the larger the absolute value of the slope, the more obvious the trend. Similarly, the linear fitting slope of the host load parameter time series data in the corresponding window is calculated as the load trend indicator. The temperature trend indicators and load trend indicators of all windows are plotted into trend curves respectively. Through comparative analysis, it is found that the change of the host load usually leads the change of the lubricating oil temperature. For example, when the host load starts to rise at time t, the lubricating oil temperature only starts to show a significant upward trend at time t+Δt, and the time difference is the potential lubricating oil-load delay.

[0054] Then, trend alignment is performed on the lubricating oil temperature trend and the host load trend with the highest trend similarity as the target. A dynamic time warping algorithm is used, with the host load trend curve as the reference sequence and the lubricating oil temperature trend curve as the sequence to be matched. By calculating the cumulative distance between the two curves, a nonlinear time warping path that minimizes the distance is found to determine the best matching position of the trend feature points in the two sequences. Based on the matching result, the delay time under different operating conditions is calculated, such as the average delay duration in the load rising stage, stable stage and falling stage. Finally, the most frequent delay duration is determined as the typical lubricating oil-load delay. Actual data verification shows that the typical delay is usually 8-12 minutes in the 300MW steam turbine generator set of the embodiment.

[0055] Afterwards, based on the determined typical lubricating oil-load delay Δt, data reorganization based on time series adjustment is performed on the lubricating oil temperature time series data and the host load parameter time series data. The host load parameter time series data is moved forward by Δt duration as a whole, so that the load data originally at time t corresponds to the lubricating oil temperature data at time t+Δt, eliminating the time series misalignment caused by the heat transfer process lag.

[0056] For example, if the host load at t=10:00 in the original data corresponds to the lubricating oil temperature at t=10:10, then after adjustment, the load data at 10:00 and the temperature data at 10:10 are combined into a set of associated samples. After performing the above adjustment on all time-series data, a dataset containing paired samples of "host load-lubricating oil temperature" is constructed, where each sample contains the adjusted host load value, current ambient temperature, lubricating oil flow rate and other auxiliary parameters, as well as the corresponding lubricating oil temperature value, establishing a delay-free mapping relationship to eliminate the influence of time delay.

[0057] Finally, based on the data reconstruction results, a load-lubricating oil temperature mapping model based on regression analysis was established. A gradient boosting regression tree was selected as the base model, which can capture nonlinear relationships and feature interactions. The host load parameters, ambient temperature, and lubricating oil circulation flow rate in the de-delayed mapping relationship were used as the model input features, and lubricating oil temperature as the output target variable. The dataset was divided into training and testing sets in an 8:2 ratio. The training set was used for model parameter learning, and the testing set was used to evaluate the model's generalization ability.

[0058] Furthermore, during model training, key hyperparameters are optimized using 5-fold cross-validation, such as setting the learning rate to 0.05, the maximum tree depth to 6, and the minimum number of samples per leaf node to 10, in order to balance the model's fitting ability and the risk of overfitting.

[0059] After training, the model performance is verified using test set data. If the root mean square error of the model is less than 0.5℃ and the mean absolute error is less than 0.3℃, the model is considered to have met the preset accuracy requirements. This load-lubricating oil temperature mapping model can be used to accurately predict the trend of lubricating oil temperature change based on the current and predicted changes in host load.

[0060] Specifically, the thermal dynamic prediction unit 13 in the system includes: Based on the real-time operating data, the real-time host load of the host system is obtained, and the planned host load for a future preset time period is obtained simultaneously. The real-time host load and the planned host load are combined and input into a pre-built load-lubricating oil temperature mapping model to obtain the predicted lubricating oil temperature sequence; Curve fitting is performed on the predicted lubricating oil temperature sequence to obtain the trend of thermal state change.

[0061] In this embodiment, firstly, the real-time main load of the main engine system at the current moment is read from the power plant's real-time data acquisition and monitoring system, for example, the actual output power of the steam turbine is currently 280MW. Simultaneously, the planned main engine load for a future preset time period is obtained from the power plant's production scheduling management system. This preset time period can be set to the next 2 hours according to actual needs. The planned load sequence may include scheduling instructions that gradually increase from the current 280MW to the rated load of 300MW and maintain operation at 300MW.

[0062] Secondly, the real-time host load is used as the starting point of the prediction sequence, and it is concatenated with the planned host load within a preset future time period in chronological order to form a complete load change sequence. For example, the time step is 5 minutes, with a total of 25 data points, including 1 current real-time point and 24 planned future points. Simultaneously, the ambient temperature corresponding to this load sequence is extracted from the real-time operating data, such as the current ambient temperature of 25℃. Auxiliary input parameters, such as assuming the ambient temperature remains stable for the next 2 hours and the real-time lubricating oil volumetric flow rate, are also input along with the load change sequence into a pre-constructed load-lubricating oil temperature mapping model. The model uses an internal gradient boosting regression tree structure to perform hierarchical feature combination and nonlinear mapping calculations on the input load and auxiliary parameters, predicting the lubricating oil temperature at each time step. Finally, it outputs a predicted lubricating oil temperature sequence of the same length as the load sequence, for example, gradually rising from the current 55℃ to 58℃ and then stabilizing.

[0063] Finally, curve fitting is performed on the output predicted lubricating oil temperature sequence. Cubic spline interpolation is used, with time as the horizontal axis and predicted lubricating oil temperature as the vertical axis, to construct a smooth fitting curve between adjacent predicted points, eliminating possible jumps in discrete data points.

[0064] Furthermore, by analyzing the slope changes, extreme point distribution, and overall trend of the fitted curve, the thermal state change trend of the lubricating oil system within a preset time period can be obtained, such as determining the starting moment of temperature rise, the stage in which the maximum heating rate occurs, and the equilibrium value after the temperature stabilizes.

[0065] The feedforward control calculation unit 14 is used to combine the real-time performance baseline with the thermal state change trend to calculate the feedforward control quantity for adjusting the flow rate of the heat transfer medium with the goal of maintaining the stable temperature in front of the lubricating oil tank. In this embodiment, maintaining the lubricating oil tank temperature at a set value is the control objective, for example, the target temperature is set to 45℃±1℃. The thermal state change trend output by the thermal dynamics prediction unit, i.e., the predicted lubricating oil temperature sequence and its fitting curve within a preset future time period, and the current real-time performance baseline of the heat exchanger, are used to calculate the required adjustment amount of the heat transfer medium flow rate to offset the predicted lubricating oil temperature fluctuations.

[0066] The composite control output unit 15 is used to generate a composite control command for the flow rate of the heat transfer medium based on the feedforward control quantity, and to perform heat recovery control.

[0067] In this embodiment, the feedforward control quantity is used as the dominant adjustment command, and a feedback correction mechanism is introduced. By comparing the deviation between the actual temperature in front of the current lubricating oil tank and the set target temperature, the feedforward control quantity is dynamically corrected.

[0068] Specifically, when the actual temperature is higher than the upper limit of the target temperature, the flow rate of the heat transfer medium is increased based on the original feedforward control quantity to enhance the heat exchange effect and accelerate the removal of heat; when the actual temperature is lower than the lower limit of the target temperature, the flow rate of the heat transfer medium is reduced to decrease the heat exchange intensity and avoid excessive drop in oil temperature.

[0069] At the same time, upper and lower limits of control quantities are set. For example, the adjustment range of the heat transfer medium flow rate is limited to 30% to 110% of the rated flow rate to prevent local overheating caused by the medium velocity in the heat exchanger being too low due to the flow rate being too small, or the flow rate being too large and exceeding the pipeline conveying capacity and pump unit operating range.

[0070] Finally, the modified and constrained composite control quantity is converted into a standard 4-20mA current signal and sent to the electric regulating valve on the heat transfer medium pipeline. The flow rate of the heat transfer medium is controlled by the continuous change of the valve opening, thereby dynamically matching the real-time thermal state changes of the lubricating oil system.

[0071] Specifically, the composite control output unit 15 in the system includes: Update and obtain the real-time lubricating oil outlet temperature of the lubrication system; Calculate the temperature control deviation between the real-time lubricating oil outlet temperature and the target set temperature in front of the lubricating oil tank; Based on the temperature control deviation and combined with the preset simplified heat transfer model, a feedback control quantity for the flow rate of the heat transfer medium is generated. The feedback control quantity and the feedforward control quantity are superimposed to generate the composite control command, and the composite control command is sent to the flow regulation mechanism of the heat transfer medium to perform heat recovery control.

[0072] In this embodiment of the application, firstly, the real-time lubricating oil outlet temperature of the lubricating oil system, i.e. the temperature in front of the lubricating oil tank, is collected and updated in real time by a temperature sensor installed on the lubricating oil outlet heat exchanger pipe. The sampling frequency is consistent with the prediction time step of the thermal dynamic prediction unit, for example, once every 5 minutes.

[0073] Secondly, calculate the temperature control deviation between the real-time lubricating oil outlet temperature and the target set temperature before the lubricating oil tank. The target set temperature is determined according to the unit's operating procedures. For example, if it is set to 45℃, then the temperature control deviation ΔT = real-time lubricating oil outlet temperature - 45℃. When ΔT is positive, it indicates that the oil temperature is too high and heat exchange needs to be enhanced; when ΔT is negative, it indicates that the oil temperature is too low and heat exchange needs to be reduced.

[0074] Then, based on the temperature control deviation ΔT and a pre-defined simplified heat transfer model, a feedback control quantity for the heat transfer medium flow rate is generated. The simplified heat transfer model is based on the heat transfer rate equation Q=K×A×ΔTm, where Q is the heat exchange, K is the current heat transfer coefficient in the real-time performance baseline, A is the heat exchanger area, and ΔTm is the logarithmic mean temperature difference. Furthermore, by inversely calculating the formula, given the target heat exchange change ΔQ, the required adjustment amount of the heat transfer medium flow rate is determined.

[0075] Subsequently, the feedback control quantity and the feedforward control quantity are superimposed to generate a composite control command. The superposition method adopts a weighted summation, with the feedforward control quantity accounting for 70% of the weight and the feedback control quantity accounting for 30% of the weight, to highlight the dominant role of the feedforward control, while the feedback control corrects the deviation in real time.

[0076] For example, the feedforward control quantity is the medium flow rate that needs to be reduced from the current 80m³ / h. 3 / h increased to 90m 3 / h, the feedback control quantity needs to be increased by an additional 5m due to ΔT=+1℃. 3 / h, then the composite control command is 90+5=95m 3 / h.

[0077] Finally, the composite control command is sent to the flow regulation mechanism of the heat transfer medium, such as an electric regulating valve. After receiving the 4-20mA standard current signal, the regulating mechanism adjusts the valve opening according to the linear correspondence between the signal and the flow rate, such as 4mA corresponding to 30% of the rated flow rate and 20mA corresponding to 110% of the rated flow rate, thereby changing the flow rate of the heat transfer medium in real time and ensuring that the temperature before the lubricating oil tank remains stable within the target range.

[0078] In summary, compared with existing technologies, this application achieves accurate prediction of lubricating oil temperature change trends by constructing a load-lubricating oil temperature mapping model based on time alignment. Combined with a feedforward-feedback composite control strategy, it can actively adjust the flow rate of the heat transfer medium before the host load changes, thereby controlling the temperature in front of the lubricating oil tank within the set range. This effectively overcomes the response lag problem caused by heat transfer delay in traditional feedback control and improves the control accuracy and dynamic response performance of the heat recovery system.

[0079] In summary, the embodiments of this application have at least the following technical effects: This application provides a heat recovery system for a power plant turbine unit's lubricating oil system. First, a data acquisition unit collects real-time operating data and historical operating logs of the lubricating oil system, heat transfer medium recovery system, and main unit under a target scenario. Second, a performance baseline update unit, based on historical operating logs, dynamically identifies the heat exchanger's performance baseline, obtaining a real-time performance baseline characterizing the current heat transfer performance and reflecting the heat exchanger's actual heat transfer capacity under different operating conditions. Third, a thermal dynamic prediction unit uses real-time operating data to predict the thermal state change trend of the lubricating oil system within a preset future time period, enabling the system to anticipate potential fluctuations in lubricating oil temperature and shift from passive control to proactive prevention. Then, a feedforward control calculation unit combines the real-time performance baseline and thermal state change trend to calculate a feedforward control quantity with the goal of maintaining a stable temperature before the lubricating oil tank, pre-adjusting the heat transfer medium flow rate from the source to reduce the possibility of temperature fluctuations. Finally, a composite control output unit generates composite control commands based on the feedforward control quantity to execute heat recovery control, considering not only predicted thermal state changes but also actual temperature deviations during operation for feedback adjustment, thereby improving heat recovery efficiency.

[0080] Through the above technical solution, this application reduces the energy consumption of the cooling equipment and effectively improves the problem of lag in lubricating oil temperature control, achieving the dual goals of efficient heat recovery and temperature control of the lubricating oil system.

[0081] Example 2, as Figure 2 As shown, based on the same inventive concept as the heat recovery system for a power plant turbine lubricating oil system provided in Embodiment 1, this application also provides a method for heat recovery in a power plant turbine lubricating oil system, including: Real-time collection of real-time operational data and historical operational logs from multiple systems in the target scenario; Based on the historical operation log, dynamic performance baseline identification is performed on the heat exchanger between the lubricating oil system and the heat transfer medium recovery system to obtain a real-time performance baseline characterizing the current heat transfer performance. Based on the real-time operating data, predict the trend of thermal state change of the lubricating oil system within a future preset time period; Based on the real-time performance baseline and the thermal state change trend, with the goal of maintaining a stable temperature in front of the lubricating oil tank, a feedforward control quantity for adjusting the flow rate of the heat transfer medium is calculated. Based on the feedforward control quantity, a composite control command for the heat transfer medium flow rate is generated to execute heat recovery control.

[0082] This includes real-time collection of operational data and historical logs from multiple systems in the target scenario, including: Obtain the lubricating oil system's operating parameters and historical lubricating oil operating logs; Obtain the operating parameters and historical operating logs of the heat transfer medium recovery system; Obtain host system load parameters and historical host operation logs; The lubricating oil operating parameters, the medium operating parameters, and the main unit load parameters are output as the real-time operating data; The historical lubricating oil operation log, the historical media operation log, and the historical host operation log are output as the historical operation log.

[0083] Furthermore, based on the historical operation logs, dynamic performance baseline identification is performed on the heat exchanger between the lubricating oil system and the heat transfer medium recovery system to obtain a real-time performance baseline characterizing the current heat transfer performance, including: Based on a preset neighborhood time window, nearby data is extracted from the historical operation log to obtain the nearby operation log; Based on the nearby operation log, the first lubricating oil temperature at the main unit outlet of the lubricating oil system, the second lubricating oil temperature in front of the lubricating oil tank, and the lubricating oil volume flow rate are obtained and output as the first real-time parameter set. Based on the nearby operation log, the temperature before recovery, the temperature after recovery, and the volumetric flow rate of the heat transfer medium recovery system are obtained and output as a second real-time parameter set. Based on the first real-time parameter set and the second real-time parameter set, and combined with a preset heat transfer basic model, an inversion operation is performed to obtain the real-time performance baseline.

[0084] Specifically, in one embodiment of the application, based on the historical operating logs, dynamic performance baseline identification is performed on the heat exchanger between the lubricating oil system and the heat transfer medium recovery system to obtain a real-time performance baseline characterizing the current heat transfer performance, including: Obtain the sequence of media-side operating parameters for the current time window; Input the sequence of operating parameters on the medium side into the pre-constructed heat transfer performance degradation mapping model to obtain the heat transfer performance degradation factor; The real-time performance baseline is calculated based on the performance degradation factor and the baseline heat transfer performance obtained from the design data.

[0085] The construction of the attenuation mapping model includes: Obtain historical media operation logs and extract the media-side operation parameter sequence, wherein the media-side operation parameter sequence includes the average flow rate and average inlet temperature of the heat transfer medium; The historical average heat transfer coefficient, which is consistent with the time sequence of the medium-side operating parameters, is obtained from the historical operation log. The historical average heat transfer coefficient is calculated based on the inversion of a preset basic heat transfer model. The heat transfer performance degradation mapping model is obtained by using the feature vector formed by the average flow rate and average inlet temperature of the heat transfer medium as the model input and the corresponding historical average heat transfer coefficient as the model output target, and supervising the training of the preset regression model.

[0086] Furthermore, based on the real-time operating data, the trend of thermal state changes in the lubricating oil system within a preset time period is predicted, including: Based on the historical operation logs, extract the lubricating oil temperature time-series data of the lubricating oil system and the host load parameter time-series data of the host system. Trend analysis was performed on the time-series data of lubricating oil temperature and the time-series data of host load parameters to obtain the trends of lubricating oil temperature and host load. With the goal of maximizing trend similarity, the lubricating oil temperature trend and the main engine load trend are trend aligned, and the typical lubricating oil-load delay is determined based on the trend alignment results; Based on the typical lubricating oil-load delay, the lubricating oil temperature time-series data and the host load parameter time-series data are reorganized based on time-series adjustment to establish a delay-free mapping relationship. Based on the data reconstruction results, a load-lubricating oil temperature mapping model based on regression analysis was established.

[0087] Among them, based on the real-time operating data, predicting the thermal state change trend of the lubricating oil system within a preset time period includes: Based on the real-time operating data, the real-time host load of the host system is obtained, and the planned host load for a future preset time period is obtained simultaneously. The real-time host load and the planned host load are combined and input into a pre-built load-lubricating oil temperature mapping model to obtain the predicted lubricating oil temperature sequence; Curve fitting is performed on the predicted lubricating oil temperature sequence to obtain the trend of thermal state change.

[0088] Further, based on the feedforward control quantity, a composite control command for the heat transfer medium flow rate is generated to execute heat recovery control, including: Update and obtain the real-time lubricating oil outlet temperature of the lubrication system; Calculate the temperature control deviation between the real-time lubricating oil outlet temperature and the target set temperature in front of the lubricating oil tank; Based on the temperature control deviation and combined with the preset simplified heat transfer model, a feedback control quantity for the flow rate of the heat transfer medium is generated. The feedback control quantity and the feedforward control quantity are superimposed to generate the composite control command, and the composite control command is sent to the flow regulation mechanism of the heat transfer medium to perform heat recovery control.

[0089] Furthermore, the lubricating oil system and the heat transfer medium recovery system exchange heat through a heat exchange component located in front of the lubricating oil tank.

[0090] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0091] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0092] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A heat recovery system for the lubricating oil system of a power plant turbine unit, characterized in that, include: The data acquisition unit is used to collect real-time operating data and historical operating logs from multiple systems in the target scenario. The performance baseline update unit is used to identify the dynamic performance baseline of the heat exchanger between the lubricating oil system and the heat transfer medium recovery system based on the historical operation log, and obtain the real-time performance baseline characterizing the current heat transfer performance. The thermal dynamic prediction unit is used to predict the trend of thermal state change of the lubricating oil system within a future preset period based on the real-time operating data. The feedforward control calculation unit is used to combine the real-time performance baseline with the thermal state change trend to calculate the feedforward control quantity for adjusting the flow rate of the heat transfer medium with the goal of maintaining the stable temperature in front of the lubricating oil tank. The composite control output unit is used to generate a composite control command for the flow rate of the heat transfer medium based on the feedforward control quantity, and to execute heat recovery control.

2. The heat recovery system for the lubricating oil system of a power plant turbine unit as described in claim 1, characterized in that, Real-time collection of real-time operational data and historical operational logs from multiple systems in the target scenario, including: Obtain the lubricating oil system's operating parameters and historical lubricating oil operating logs; Obtain the operating parameters and historical operating logs of the heat transfer medium recovery system; Obtain host system load parameters and historical host operation logs; The lubricating oil operating parameters, the medium operating parameters, and the main unit load parameters are output as the real-time operating data; The historical lubricating oil operation log, the historical media operation log, and the historical host operation log are output as the historical operation log.

3. The heat recovery system for the lubricating oil system of a power plant turbine unit as described in claim 1, characterized in that, Based on the historical operation logs, dynamic performance baseline identification is performed on the heat exchanger between the lubricating oil system and the heat transfer medium recovery system to obtain a real-time performance baseline characterizing the current heat transfer performance, including: Based on a preset neighborhood time window, nearby data is extracted from the historical operation log to obtain the nearby operation log; Based on the nearby operation log, the first lubricating oil temperature at the main unit outlet of the lubricating oil system, the second lubricating oil temperature in front of the lubricating oil tank, and the lubricating oil volume flow rate are obtained and output as the first real-time parameter set. Based on the nearby operation log, the temperature before recovery, the temperature after recovery, and the volumetric flow rate of the heat transfer medium recovery system are obtained and output as a second real-time parameter set. Based on the first real-time parameter set and the second real-time parameter set, and combined with a preset heat transfer basic model, an inversion operation is performed to obtain the real-time performance baseline.

4. The heat recovery system for the lubricating oil system of a power plant turbine unit as described in claim 1, characterized in that, Based on the historical operation logs, dynamic performance baseline identification is performed on the heat exchanger between the lubricating oil system and the heat transfer medium recovery system to obtain a real-time performance baseline characterizing the current heat transfer performance, including: Obtain the sequence of media-side operating parameters for the current time window; Input the sequence of operating parameters on the medium side into the pre-constructed heat transfer performance degradation mapping model to obtain the heat transfer performance degradation factor; The real-time performance baseline is calculated based on the performance degradation factor and the baseline heat transfer performance obtained from the design data.

5. A heat recovery system for a power plant turbine unit lubricating oil system as described in claim 4, characterized in that, The construction of the attenuation mapping model includes: Obtain historical media operation logs and extract the media-side operation parameter sequence, wherein the media-side operation parameter sequence includes the average flow rate and average inlet temperature of the heat transfer medium; The historical average heat transfer coefficient, which is consistent with the time sequence of the medium-side operating parameters, is obtained from the historical operation log. The historical average heat transfer coefficient is calculated based on the inversion of a preset basic heat transfer model. The heat transfer performance degradation mapping model is obtained by using the feature vector formed by the average flow rate and average inlet temperature of the heat transfer medium as the model input and the corresponding historical average heat transfer coefficient as the model output target, and supervising the training of the preset regression model.

6. A heat recovery system for a power plant turbine unit lubricating oil system as described in claim 1, characterized in that, Based on the real-time operating data, predict the trend of thermal state changes in the lubricating oil system within a preset time period, including: Based on the historical operation logs, extract the lubricating oil temperature time-series data of the lubricating oil system and the host load parameter time-series data of the host system. Trend analysis was performed on the time-series data of lubricating oil temperature and the time-series data of host load parameters to obtain the trends of lubricating oil temperature and host load. With the goal of maximizing trend similarity, the lubricating oil temperature trend and the main engine load trend are trend aligned, and the typical lubricating oil-load delay is determined based on the trend alignment results; Based on the typical lubricating oil-load delay, the lubricating oil temperature time-series data and the host load parameter time-series data are reorganized based on time-series adjustment to establish a delay-free mapping relationship. Based on the data reconstruction results, a load-lubricating oil temperature mapping model based on regression analysis was established.

7. A heat recovery system for a power plant turbine unit lubricating oil system as described in claim 1, characterized in that, Based on the real-time operating data, predict the trend of thermal state changes in the lubricating oil system within a preset time period, including: Based on the real-time operating data, the real-time host load of the host system is obtained, and the planned host load for a future preset time period is obtained simultaneously. The real-time host load and the planned host load are combined and input into a pre-built load-lubricating oil temperature mapping model to obtain the predicted lubricating oil temperature sequence; Curve fitting is performed on the predicted lubricating oil temperature sequence to obtain the trend of thermal state change.

8. A heat recovery system for a power plant turbine unit lubricating oil system as described in claim 1, characterized in that, Based on the feedforward control quantity, a composite control command for the heat transfer medium flow rate is generated to execute heat recovery control, including: Update and obtain the real-time lubricating oil outlet temperature of the lubrication system; Calculate the temperature control deviation between the real-time lubricating oil outlet temperature and the target set temperature in front of the lubricating oil tank; Based on the temperature control deviation and combined with the preset simplified heat transfer model, a feedback control quantity for the flow rate of the heat transfer medium is generated. The feedback control quantity and the feedforward control quantity are superimposed to generate the composite control command, and the composite control command is sent to the flow regulation mechanism of the heat transfer medium to perform heat recovery control.

9. A heat recovery system for a power plant turbine unit lubricating oil system as described in claim 1, characterized in that, The lubricating oil system and the heat transfer medium recovery system exchange heat through a heat exchange component located in front of the lubricating oil tank.

10. A method for heat recovery in the lubricating oil system of a power plant turbine unit, characterized in that, A heat recovery system for a power plant turbine unit lubricating oil system according to any one of claims 1-9, comprising: Real-time collection of real-time operational data and historical operational logs from multiple systems in the target scenario; Based on the historical operation log, dynamic performance baseline identification is performed on the heat exchanger between the lubricating oil system and the heat transfer medium recovery system to obtain a real-time performance baseline characterizing the current heat transfer performance. Based on the real-time operating data, predict the trend of thermal state change of the lubricating oil system within a future preset time period; Based on the real-time performance baseline and the thermal state change trend, with the goal of maintaining a stable temperature in front of the lubricating oil tank, a feedforward control quantity for adjusting the flow rate of the heat transfer medium is calculated. Based on the feedforward control quantity, a composite control command for the heat transfer medium flow rate is generated to execute heat recovery control.

Images