A boiler full-process collaborative intelligent control system

Abstract

The present invention discloses a boiler full-process collaborative intelligent control system, which relates to the field of boiler intelligent control, including a data processing module, which acquires multi-dimensional data of boiler operation in real time and performs preprocessing, a feature fusion module, which extracts and fuses features of the preprocessed data to obtain key feature vectors, a boiler prediction module, which establishes a boiler dynamic prediction model based on the key feature vectors and outputs the future operating state of the boiler, and an optimization target module, which establishes a boiler optimization model based on the future operating state of the boiler and generates a boiler optimization target. The present invention can achieve a dynamic balance between environmental protection, energy saving and efficiency through the design of a boiler optimization model, and adjust the optimization weights according to different scenarios to achieve multi-objective collaborative optimization. At the same time, user preferences are introduced to make the optimization target more targeted.

Description

Technical Field

[0001] The present invention relates to the field of boiler intelligent control, and in particular to a boiler full-process collaborative intelligent control system. Background Art

[0002] As key equipment in industrial production and energy supply, boilers are widely used in power, chemical, metallurgy, construction and other fields, and have a significant impact on energy efficiency and environmental emission indicators. With the rapid development of industrial automation and intelligent technology, traditional boiler control systems are gradually transforming towards intelligent and data-driven directions.

[0003] Existing boiler intelligent control technologies lack user preference guidance and only focus on boiler equipment performance indicators, but fail to combine the specific needs of user application scenarios. In addition, existing boiler optimization models are mostly based on a single objective, making it difficult to achieve global optimization in multi-objective optimization. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a boiler full-process collaborative intelligent control system to solve the problem of lack of user preference guidance, only focusing on the equipment performance indicators of the boiler, and difficulty in achieving global optimization in multi-objective optimization.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] The present invention provides a boiler full-process collaborative intelligent control system, which includes:

[0008] Data processing module, which acquires multi-dimensional data of boiler operation in real time and performs pre-processing;

[0009] The feature fusion module extracts and fuses the features of the preprocessed data to obtain the key feature vector;

[0010] The boiler prediction module establishes a boiler dynamic prediction model based on key feature vectors and outputs the future operating status of the boiler;

[0011] The optimization target module establishes a boiler optimization model based on the future operating status of the boiler and generates the boiler optimization target;

[0012] User preference module, which introduces user preferences and forms user preference weights;

[0013] The operating parameter module calculates the optimal operating parameters of the boiler based on the boiler user's preferred optimization target and the current operating status of the boiler;

[0014] The detection module executes the optimal operating parameters of the boiler and continuously detects the operating status of the boiler;

[0015] The correction module dynamically corrects the operating parameters according to the actual operating status of the boiler.

[0016] As a preferred solution of the boiler full-process collaborative intelligent control system of the present invention, the multi-dimensional data of boiler operation is obtained in real time and pre-processed, which specifically includes the following steps:

[0017] Real-time collection of boiler operation fuel ratio, calorific value of each fuel, oxygen content, exhaust gas temperature, steam flow, maximum steam flow, ambient temperature and emission parameters;

[0018] The threshold filtering method is used to remove outliers in the data, and the linear interpolation method is used to fill in the missing data points.

[0019] As a preferred solution of the boiler full-process collaborative intelligent control system of the present invention, the following steps are specifically included: feature extraction and fusion of the pre-processed data to obtain key feature vectors:

[0020] According to the fuel ratio and the calorific value of each fuel, the calorific value of the fuel is obtained;

[0021] Calculate the ratio of the current boiler load to the maximum design load based on the steam flow rate and the maximum steam flow rate;

[0022] Calculate the excess air coefficient required for combustion through the amount of oxygen;

[0023] Calculate the pollutant emissions per unit load based on emission parameters;

[0024] Calculate heat loss based on exhaust gas temperature and ambient temperature;

[0025] The fuel calorific value, boiler load factor, excess air coefficient, pollutant emissions and heat loss are combined to form the key feature vector.

[0026] As a preferred solution of the boiler full-process collaborative intelligent control system of the present invention, a boiler dynamic prediction model is established according to key feature vectors to output the future operating state of the boiler, which specifically includes the following steps:

[0027] The static features of the key feature vector are extracted using a nonlinear transformation function, and its expression is:

[0028]

[0029] Among them, φ(X t ) represents the static characteristics of the current key feature vector of the boiler, X t represents the key feature vector, t represents the time point, X irepresents the i-th sub-feature in the key feature vector, n represents the dimension of the feature vector, and sigmoid() represents the nonlinear activation function used for normalization;

[0030] The time gradient function is used to extract the dynamic change characteristics of the key feature vector, and its expression is:

[0031]

[0032] in, Represents the dynamic change characteristics of key feature vectors;

[0033] Weighting the static features and dynamically changing features of the key feature vector;

[0034] The weighted static features and dynamic change features are subjected to nonlinear mapping of hyperbolic tangent, and the operating status of the boiler in the future time span is output by accumulating the predicted values. The expression is:

[0035]

[0036] in, It represents the predicted operating status of the boiler at the future time t+Δt, t represents the current time, Δt represents the predicted time span, tanh() represents the hyperbolic tangent, W1 represents the weight parameter for static features, W2 represents the weight parameter for dynamically changing features, b represents the bias parameter, and dt represents a small change in the time variable.

[0037] As a preferred solution of the boiler full-process collaborative intelligent control system of the present invention, a boiler optimization model is established based on the future operating state of the boiler to generate a boiler optimization target, which specifically includes the following steps:

[0038] According to the predicted future operating state of the boiler, the boiler optimization target is generated by optimizing the objective function, and its expression is:

[0039]

[0040] in, represents the boiler optimization target, E(t) represents the boiler pollutant emission at the current time, P(t) represents the boiler energy consumption at the current time, L(t) represents the boiler heat loss at the current time, α represents the weight of the boiler pollutant emission, β represents the weight of the boiler energy consumption, γ represents the weight of the boiler heat loss, and min represents the minimum value calculation.

[0041] As a preferred solution of the boiler full-process collaborative intelligent control system of the present invention, the following steps are specifically included: introducing user preferences and forming user preference weights:

[0042] Introduce user preference function to collect user preferences in different scenarios;

[0043] According to the real-time operating status of the boiler, the operating feedback factor is obtained;

[0044] Combine user preferences with operational feedback factors to dynamically calculate the weights of boiler pollutant emissions, boiler energy consumption, and boiler heat loss to form user preference weights;

[0045] The user preference weights are introduced into the boiler optimization model to output the optimization target that meets the user's current needs.

[0046] As a preferred solution of the boiler full-process collaborative intelligent control system of the present invention, wherein: the operation feedback factor is obtained, specifically comprising the following steps:

[0047] Calculate the pollutant emissions per unit time based on the steam flow rate;

[0048] Calculate the total energy consumption of the boiler per unit time based on the calorific value of the fuel;

[0049] Calculate the total losses in boiler operation based on exhaust gas temperature, ambient temperature and boiler heat dissipation;

[0050] The feedback factors of boiler pollutant emissions, total energy consumption and total losses in operation are calculated through normalization.

[0051] As a preferred solution of the boiler full-process collaborative intelligent control system of the present invention, wherein: according to the boiler user preference optimization target and the current operating state of the boiler, the optimal operating parameters of the boiler are calculated, which specifically includes the following steps:

[0052] Set the parameters for the convergence conditions of the optimal operating parameters of the boiler;

[0053] According to the current operating parameters of the boiler, the gradient of the boiler user's preferred optimization target to the current operating parameters is calculated;

[0054] Use the results of gradient calculation to update the current operating parameters of the boiler;

[0055] Determine whether the current boiler parameters meet the convergence conditions. When the current boiler parameters meet the convergence conditions, stop the iteration and output the optimal operating parameters of the boiler.

[0056] As a preferred solution of the boiler full-process collaborative intelligent control system of the present invention, wherein: executing the optimal operating parameters of the boiler and continuously detecting the operating status of the boiler specifically includes the following steps:

[0057] The calculated optimal operating parameters of the boiler are transmitted to the boiler control unit;

[0058] The boiler control unit adjusts the operating state of the boiler according to the received optimal operating parameters of the boiler and enters the optimal operating mode of the boiler;

[0059] The boiler operation control unit monitors the feedback status of the boiler in real time to ensure that the boiler enters the optimal operation mode.

[0060] As a preferred solution of the boiler full-process collaborative intelligent control system of the present invention, the operating parameters are dynamically corrected according to the actual operating state of the boiler, specifically including the following steps:

[0061] Detect whether there is any deviation based on the feedback status of the boiler;

[0062] Dynamically correct boiler operating parameters based on deviations;

[0063] After the boiler operating parameters are corrected, the actual operating status of the boiler is monitored in real time to verify the correction effect until the deviation is eliminated.

[0064] The beneficial effects of the present invention are as follows: through the design of the boiler optimization model, a dynamic balance can be achieved between environmental protection, energy saving and efficiency, and the optimization weights can be adjusted according to different scenarios to achieve multi-objective collaborative optimization. At the same time, user preferences are introduced to support specific optimization goals of different users, and more expected optimization solutions are provided according to user needs, making the optimization goals more targeted. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0066] Figure 1 This is a module diagram of the boiler full-process collaborative intelligent control system in Example 1.

[0067] Figure 2 This is a schematic diagram of the boiler optimization model in Example 1. DETAILED DESCRIPTION

[0068] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0069] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0070] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0071] Example 1, reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, and provides a boiler full-process collaborative intelligent control system, including the following steps:

[0072] Data processing module, which acquires multi-dimensional data of boiler operation in real time and performs pre-processing;

[0073] The fuel ratio, calorific value of each fuel, oxygen content, exhaust temperature, steam flow, maximum steam flow, ambient temperature and emission parameters of the boiler are collected in real time through sensors, and the collection frequency is set to once per second.

[0074] The data is collected by the sensor and transmitted to the data acquisition server, and is also passed to the data processing unit through the interface.

[0075] Through the threshold filtering method, a reasonable range is defined for each variable, and values ​​outside the range are eliminated.

[0076] Taking oxygen content as an example, the reasonable range is greater than or equal to 2% and less than or equal to 10%. If it does not meet the reasonable range, it will be eliminated.

[0077] For missing data points, linear interpolation is used to fill them in. The missing data is generated through the linear trend of the two previous and next data points to avoid affecting the subsequent feature extraction.

[0078] The feature fusion module extracts and fuses the features of the preprocessed data to obtain the key feature vector;

[0079] The weighted average calorific value is calculated based on the fuel ratio and the calorific value of each fuel to obtain the fuel calorific value.

[0080] Calculate the ratio of the current boiler load to the maximum design load to obtain the boiler load rate.

[0081] Specifically, boiler load factor = steam flow rate / maximum steam flow rate.

[0082] The excess air ratio required for combustion is calculated based on the amount of oxygen.

[0083] Specifically, excess air ratio=(0.2 / 21)×(1 / theoretical air ratio).

[0084] Calculate the pollutant emissions under unit load and obtain the pollutant emission intensity.

[0085] Specifically, NOx intensity=NOx concentration / steam flow rate.

[0086] To further explain, emission intensity is used to evaluate environmental performance.

[0087] Calculate heat loss based on exhaust gas temperature and ambient temperature.

[0088] Specifically, exhaust heat loss = smoke specific heat capacity × (exhaust temperature - ambient temperature).

[0089] The calculated fuel calorific value, boiler load rate, excess air coefficient, pollutant emission intensity and exhaust heat loss are integrated to form the key feature vector X t .

[0090] The boiler prediction module establishes a boiler dynamic prediction model based on key feature vectors and outputs the future operating status of the boiler;

[0091] The static features of the key feature vector are extracted using a nonlinear transformation function to capture the overall operating characteristics of the boiler at the current moment. The expression is:

[0092]

[0093] Among them, φ(X t ) represents the static characteristics of the current key feature vector of the boiler, X t represents the key feature vector, t represents the time point, X i represents the i-th sub-feature in the key feature vector, n represents the dimension of the feature vector, and sigmoid() represents the nonlinear activation function used for normalization;

[0094] It is further explained that static feature extraction normalizes the current features to eliminate the influence of different feature dimensions, and uses the sigmoid activation function to smooth the feature values ​​and enhance the nonlinear fitting ability of the model.

[0095] The dynamic characteristics reflect the changing trend of the boiler operating status over time and describe the impact of the changing rate of operating parameters on the boiler.

[0096] First, calculate the rate of change of the sub-feature over time, and then multiply the rate of change by ln(1+X i ) to enhance the nonlinear characteristics of feature changes, and finally add the results of all sub-features to obtain the dynamic eigenvalue. Its expression is:

[0097]

[0098] in, Represents the dynamic change characteristics of key feature vectors;

[0099] After weighting the static and dynamic features, the operating status of the boiler at the future moment is output through nonlinear mapping, and its expression is:

[0100]

[0101] in, It represents the predicted operating status of the boiler at the future time t+Δt, t represents the current time, Δt represents the predicted time span, tanh() represents the hyperbolic tangent, W1 represents the weight parameter for static features, W2 represents the weight parameter for dynamically changing features, b represents the bias parameter, and dt represents a small change in the time variable.

[0102] The result is mapped to the interval (0,1) to indicate the operating status of the boiler. When the output is close to 1, it means that the boiler is operating well. When the output is close to -1, it means that the boiler is operating abnormally.

[0103] By accumulating the predicted values ​​in the future time period, the operating status of the boiler in the future time span is obtained.

[0104] It is further explained that the advantages of static and dynamic features are combined, so that the model can simultaneously capture the current operating status of the boiler and future change trends, and make predictions based on the feature vector and its change rate at the current moment, thus realizing real-time status assessment.

[0105] The optimization target module establishes a boiler optimization model based on the future operating status of the boiler and generates the boiler optimization target;

[0106] Calculate the comprehensive operating cost of the boiler in the future time period,

[0107] Specifically, it is obtained by accumulating the pollutant emissions, boiler energy consumption and boiler heat loss by weight.

[0108] Pollutant emissions are used to describe the degree of environmental pollution generated during boiler operation.

[0109] For example, sulfur dioxide (SO2) and nitrogen oxides (NO x ) and particulate matter (PM), etc., are the focus of environmental protection.

[0110] Boiler energy consumption is used to reflect the energy consumption required for boiler operation. Energy consumption directly affects the operating cost of the boiler. Reducing energy consumption is the core goal of improving boiler efficiency.

[0111] Boiler heat loss reflects the heat energy loss caused by heat dissipation from the furnace wall, smoke exhaust, incomplete combustion, etc. during the operation of the boiler. Heat loss directly affects the thermal efficiency of the boiler. Reducing heat loss is an important direction for boiler optimization.

[0112] According to the accumulated value of the predicted future operating state of the boiler within the time interval, the operating stability of the boiler in the future time period is obtained.

[0113] The comprehensive operating cost of the boiler in the future time period and the operating stability of the boiler in the future time period are integrated to form the boiler optimization target, which is expressed as:

[0114]

[0115] in, represents the boiler optimization target, E(t) represents the boiler pollutant emission at the current time, P(t) represents the boiler energy consumption at the current time, L(t) represents the boiler heat loss at the current time, α represents the weight of the boiler pollutant emission, β represents the weight of the boiler energy consumption, γ represents the weight of the boiler heat loss, and min represents the minimum value calculation.

[0116] Optimization target value It is a dimensionless value that represents the balance between the overall efficiency and cost of boiler operation.

[0117] It is further explained that multi-objective optimization can be achieved through the weighted combination of the three key performance indicators of pollutant emissions, energy consumption and heat loss. The realization of the optimization goals can significantly reduce the operating costs of the boiler, including fuel costs, maintenance costs and environmental protection costs, and reduce resource waste while ensuring efficient operation of the boiler.

[0118] User preference module, which introduces user preferences and forms user preference weights;

[0119] Through the user preference function, the user's needs in different scenarios are collected.

[0120] For example: some users pay more attention to environmental protection (reducing pollutant emissions); some users pay more attention to costs (reducing energy consumption); some users give priority to boiler efficiency (reducing heat loss).

[0121] The pollutant emissions per unit time are calculated based on the steam flow rate, and the expression is:

[0122] E(t) = Q × C × η;

[0123] Among them, E(t) represents the pollutant emission of the boiler per unit time, Q represents the steam flow rate of the boiler per unit time (tons / hour), C represents the content of pollutants in the fuel (such as sulfur content, nitrogen content, etc.), the unit is kg / ton, and η represents the boiler combustion efficiency, which represents the proportion of the fuel actually participating in the reaction during the combustion process.

[0124] The total energy consumption per unit time of the boiler is calculated based on the calorific value of the fuel:

[0125] P(t) = A × H × η;

[0126] Among them, P(t) represents the total energy consumption of the boiler per unit time (kWh), A represents the fuel consumption (tons / hour), and H represents the lower calorific value of the fuel (kJ / kg).

[0127] The total loss in boiler operation is calculated based on the exhaust gas temperature, ambient temperature and boiler heat dissipation, and the expression is:

[0128] L(t)=Q×(T ex -T env )×J+B;

[0129] Where L(t) represents the total heat loss of the boiler per unit time, T ex Indicates boiler exhaust temperature, T env represents the ambient temperature, J represents the specific heat capacity of the flue gas, and B represents the heat dissipated by the boiler wall.

[0130] The pollutant emissions, total energy consumption and total losses are normalized into the operation feedback factor, which is expressed as:

[0131]

[0132] Among them, s E (t) indicates the importance of the operating cost of the boiler’s pollutant emissions. A larger value indicates a higher cost. P (t) indicates the importance of the total energy consumption cost of the boiler. The larger the value, the higher the cost. L (t) indicates the importance of the total heat loss operating cost of the boiler. The larger the value, the higher the cost. max Indicates the maximum permissible value of pollutant emissions from the boiler, P max Indicates the maximum allowable value of the total energy consumption of the boiler, L max Indicates the maximum allowable value of the total heat loss of the boiler.

[0133] Combining user preferences and operation feedback factors, the weights in the boiler optimization objective function are dynamically adjusted, and the expression is:

[0134] α(t)=u E (t)×s E (t);

[0135] β(t)=u P (t)×s P (t);

[0136] γ(t)=u L (t)×s L (t);

[0137] Among them, α(t) represents the dynamic weight of boiler pollutant emissions, u E (t) represents the user's weight on boiler pollutant emissions, β(t) represents the dynamic weight of boiler energy consumption, and u P (t) represents the user's weight on boiler energy consumption, γ(t) represents the dynamic weight of boiler heat loss, and u L (t) represents the user's weight on boiler heat loss.

[0138] Renormalize the calculated dynamic weights to form user preference weights

[0139] The calculated user preference weights are introduced into the boiler optimization model. According to the user preferences and the real-time operation status of the boiler, the weights in the optimization objective function are dynamically adjusted to output the optimization target that meets the user's current needs. The expression is:

[0140]

[0141] in, represents the weight of boiler pollutant emissions after dynamic adjustment of user preferences, represents the boiler energy consumption weight after dynamic adjustment of user preferences, Represents the boiler heat loss weight after dynamic adjustment by user preference.

[0142] Further explanation: By collecting user preferences, the boiler optimization model can dynamically adjust the optimization target according to the user's needs in different scenarios (such as environmental protection priority, cost priority, efficiency priority). At the same time, the operation feedback factor is used to reflect the real-time operation performance of the boiler to ensure that the optimization target can adapt to the dynamic changes in the boiler operating conditions.

[0143] The operating parameter module calculates the optimal operating parameters of the boiler based on the boiler user's preferred optimization target and the current operating status of the boiler;

[0144] Set the parameters for the convergence conditions.

[0145] Specifically, if the value of the optimization target (such as comprehensive cost or environmental protection index) changes very little in two consecutive calculations, it is considered that the target is close enough to the final result and the calculation is stopped.

[0146] If the operating parameters of the boiler (such as fuel flow or air supply) change very little between two consecutive adjustments, it means that the parameters are close to the optimal value and the calculation is stopped.

[0147] Set the maximum number of calculations required for convergence.

[0148] It is important to clarify how to adjust the boiler's operating parameters to bring the target closer to user needs.

[0149] Specifically, with regard to the adjustment direction of the fuel flow rate, if the pollutant emissions increase significantly after the fuel flow rate increases, but the energy consumption does not change much, the optimization algorithm recommends reducing the fuel flow rate.

[0150] The adjustment direction of the air supply: if the air supply is insufficient, the combustion efficiency will be reduced, resulting in increased energy consumption and pollutant emissions. The optimization algorithm recommends increasing the air supply.

[0151] The adjustment direction of the boiler load rate. If the load rate is too high, the exhaust gas temperature will increase and the heat loss will increase significantly. The optimization algorithm will recommend reducing the load rate to reduce heat loss.

[0152] According to the influence of the current operating parameters on the optimization target, adjust the operating parameters of the boiler to make it closer to the optimization target. Do not adjust too much each time you update, so as to avoid fluctuations in the optimization target due to rapid parameter changes (the adjustment ratio is usually between 0.1%-5%).

[0153] After each parameter adjustment, check whether the previously set convergence conditions are met. If they are met, the calculation can be stopped; if not, continue to adjust the parameters.

[0154] Once the convergence conditions are met, the current operating parameters are considered to be the optimal parameters. These parameters are output and applied to the actual operation of the boiler.

[0155] For example, fuel flow rate: 580kg / h, air supply: 1200m 3 / h, boiler load rate: 85%, comprehensive operating cost: 1.65 yuan / ton of steam.

[0156] It is further explained that through multiple rounds of calculations and parameter iterations, the optimal operating parameters of the boiler are finally found, which can improve combustion efficiency and reduce heat loss. The operating parameters of the boiler are rationally optimized to reduce fuel consumption and energy waste, reduce operating costs, and optimize the automated design of the model, which greatly reduces the frequency of manual adjustment of boiler parameters and improves the level of intelligence in industrial production. This optimization method is applicable to different types of boilers (such as coal-fired boilers, gas-fired boilers, and biomass boilers), and can dynamically adjust parameters according to real-time operating status.

[0157] The detection module executes the optimal operating parameters of the boiler and continuously detects the operating status of the boiler;

[0158] The optimal operating parameters are transmitted to the control unit via wireless communication (such as Wi-Fi, LoRa or 5G industrial network).

[0159] After receiving the optimal operating parameters, the control unit adjusts the operating status of the boiler according to these parameters.

[0160] The adjustments include adjusting the fuel supply rate according to the optimal operating parameters; adjusting the air supply according to the optimal operating parameters to ensure sufficient combustion of the fuel; adjusting the steam or hot water output of the boiler according to user needs and the calculated boiler load rate; reducing heat losses by controlling the boiler's exhaust temperature; and ensuring that the boiler water circulation or steam flow meets the optimal operating parameters.

[0161] After entering the optimal operating mode, the boiler control unit needs to monitor the boiler's feedback status in real time to ensure that the operating parameters are consistent with the actual operating results.

[0162] The key indicators monitored are pollutant emissions, combustion efficiency, boiler heat loss, energy consumption and boiler load output.

[0163] Further explanation: By combining the optimal operating parameters with the boiler control unit, the boiler can adjust the operating status in real time and always maintain the optimal operating mode. This dynamic control method can adapt to load fluctuations, fuel changes and other working conditions to ensure efficient operation of the boiler. In addition, the real-time monitoring of the boiler's feedback status can quickly detect abnormal conditions in operation and automatically adjust, reducing the occurrence of failures and improving the reliability and stability of boiler operation.

[0164] The correction module dynamically corrects the operating parameters according to the actual operating status of the boiler.

[0165] The collected feedback data is compared with the optimization target value in real time. If the deviation exceeds the preset tolerance range, it is considered that there is a deviation;

[0166] For example, the tolerance range is ±5% or ±2°C.

[0167] Analyze the specific source of the deviation based on the abnormal data of the feedback status. For example, if the combustion efficiency is reduced, it may be caused by insufficient air supply or fluctuations in fuel quality. If the exhaust temperature is too high, it may be caused by increased heat loss or excessive fuel.

[0168] Adjust relevant operating parameters according to the specific cause of the deviation.

[0169] For example: Adjust the fuel flow: If the combustion efficiency is too low, reduce the fuel flow to avoid incomplete combustion; if the load demand increases, increase the fuel flow appropriately.

[0170] Adjust the air supply: if the oxygen content is insufficient, increase the blower output; if the oxygen is excessive, reduce the air supply.

[0171] Adjust boiler load: adjust the output flow of steam or hot water according to fluctuations in user demand.

[0172] Adjust the exhaust temperature: reduce the exhaust temperature by adjusting the flue gas recirculation device or waste heat recovery equipment.

[0173] After the operating parameters are corrected, it is necessary to monitor the actual operating status of the boiler in real time to verify whether the correction is effective. If the deviation still exists, continue to adjust the operating parameters until the deviation is completely eliminated. If there is still a deviation after the correction, further adjust the operating parameters according to the new feedback status, forming a dynamic closed loop of adjustment-verification-re-adjustment until the deviation is completely eliminated.

[0174] It is further explained that through deviation detection and dynamic correction, the boiler can quickly adapt to changes in the external environment, fuel quality and load demand, and always maintain the optimal operating mode. Real-time monitoring and dynamic correction avoid problems such as equipment overload and incomplete combustion caused by long-term deviation of operating parameters from the optimal value, and improve the operating stability and equipment life of the boiler.

[0175] In summary, the present invention can achieve a dynamic balance between environmental protection, energy saving and efficiency through the design of the boiler optimization model, and adjust the optimization weight according to different scenarios to achieve multi-objective collaborative optimization. At the same time, user preferences are introduced to support specific optimization goals of different users, and a more expected optimization solution is provided according to user needs, making the optimization goal more targeted.

[0176] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A boiler full-process collaborative intelligent control system, characterized by: include, Data processing module, which acquires multi-dimensional data of boiler operation in real time and performs pre-processing; The feature fusion module extracts and fuses the features of the preprocessed data to obtain the key feature vector; The boiler prediction module establishes a boiler dynamic prediction model based on key feature vectors and outputs the future operating status of the boiler; The optimization target module establishes a boiler optimization model based on the future operating status of the boiler and generates the boiler optimization target; User preference module, which introduces user preferences and forms user preference weights; The operating parameter module calculates the optimal operating parameters of the boiler based on the boiler user's preferred optimization target and the current operating status of the boiler; The detection module executes the optimal operating parameters of the boiler and continuously detects the operating status of the boiler; The correction module dynamically corrects the operating parameters according to the actual operating status of the boiler.

2. The boiler full-process collaborative intelligent control system according to claim 1, characterized in that: Real-time acquisition of multi-dimensional data of boiler operation and pre-processing, including the following steps: Real-time collection of boiler operation fuel ratio, calorific value of each fuel, oxygen content, exhaust gas temperature, steam flow, maximum steam flow, ambient temperature and emission parameters; The threshold filtering method is used to remove outliers in the data, and the linear interpolation method is used to fill in the missing data points.

3. The boiler full-process collaborative intelligent control system according to claim 2, characterized in that: The preprocessed data is subjected to feature extraction and fusion to obtain key feature vectors, which specifically includes the following steps: According to the fuel ratio and the calorific value of each fuel, the calorific value of the fuel is obtained; Calculate the ratio of the current boiler load to the maximum design load based on the steam flow rate and the maximum steam flow rate; Calculate the excess air coefficient required for combustion through the amount of oxygen; Calculate the pollutant emissions per unit load based on emission parameters; Calculate heat loss based on exhaust gas temperature and ambient temperature; The fuel calorific value, boiler load factor, excess air coefficient, pollutant emissions and heat loss are combined to form the key feature vector.

4. The boiler full-process collaborative intelligent control system according to claim 3, characterized in that: A boiler dynamic prediction model is established based on the key feature vectors to output the future operating status of the boiler. The specific steps include: The static features of the key feature vector are extracted using a nonlinear transformation function, and its expression is: Among them, φ(X t ) represents the static characteristics of the current key feature vector of the boiler, X t represents the key feature vector, t represents the time point, X i represents the i-th sub-feature in the key feature vector, n represents the dimension of the feature vector, and sigmoid() represents the nonlinear activation function used for normalization; The time gradient function is used to extract the dynamic change characteristics of the key feature vector, and its expression is: in, Represents the dynamic change characteristics of key feature vectors; Weighting the static features and dynamically changing features of the key feature vector; The weighted static features and dynamic change features are subjected to nonlinear mapping of hyperbolic tangent, and the operating status of the boiler in the future time span is output by accumulating the predicted values. The expression is: in, It represents the predicted operating status of the boiler at the future time t+Δt, t represents the current time, Δt represents the predicted time span, tanh() represents the hyperbolic tangent, W1 represents the weight parameter for static features, W2 represents the weight parameter for dynamically changing features, b represents the bias parameter, and dt represents a small change in the time variable.

5. The boiler full-process collaborative intelligent control system according to claim 4, characterized in that: Based on the future operating status of the boiler, a boiler optimization model is established to generate the boiler optimization target, which specifically includes the following steps: According to the predicted future operating state of the boiler, the boiler optimization target is generated by optimizing the objective function, and its expression is: in, represents the boiler optimization target, E(t) represents the boiler pollutant emission at the current time, P(t) represents the boiler energy consumption at the current time, L(t) represents the boiler heat loss at the current time, α represents the weight of the boiler pollutant emission, β represents the weight of the boiler energy consumption, γ represents the weight of the boiler heat loss, and min represents the minimum value calculation.

6. The boiler full-process collaborative intelligent control system according to claim 5, characterized in that: Introducing user preferences and forming user preference weights specifically includes the following steps: Introduce user preference function to collect user preferences in different scenarios; According to the real-time operating status of the boiler, the operating feedback factor is obtained; Combine user preferences with operational feedback factors to dynamically calculate the weights of boiler pollutant emissions, boiler energy consumption, and boiler heat loss to form user preference weights; The user preference weights are introduced into the boiler optimization model to output the optimization target that meets the user's current needs.

7. The boiler full-process collaborative intelligent control system according to claim 6, characterized in that: The obtaining of the operation feedback factor specifically comprises the following steps: Calculate the pollutant emissions per unit time based on the steam flow rate; Calculate the total energy consumption of the boiler per unit time based on the calorific value of the fuel; Calculate the total losses in boiler operation based on exhaust gas temperature, ambient temperature and boiler heat dissipation; The feedback factors of boiler pollutant emissions, total energy consumption and total losses in operation are calculated through normalization.

8. The boiler full-process collaborative intelligent control system according to claim 7, characterized in that: The optimal operating parameters of the boiler are obtained by calculating according to the optimization target preferred by the boiler user and the current operating status of the boiler. The specific steps include: Set the parameters for the convergence conditions of the optimal operating parameters of the boiler; According to the current operating parameters of the boiler, the gradient of the boiler user's preferred optimization target to the current operating parameters is calculated; Use the results of gradient calculation to update the current operating parameters of the boiler; Determine whether the current boiler parameters meet the convergence conditions. When the current boiler parameters meet the convergence conditions, stop the iteration and output the optimal operating parameters of the boiler.

9. The boiler full-process collaborative intelligent control system according to claim 8, characterized in that: Execute the optimal operating parameters of the boiler and continuously detect the operating status of the boiler, which specifically includes the following steps: The calculated optimal operating parameters of the boiler are transmitted to the boiler control unit; The boiler control unit adjusts the operating state of the boiler according to the received optimal operating parameters of the boiler and enters the optimal operating mode of the boiler; The boiler operation control unit monitors the feedback status of the boiler in real time to ensure that the boiler enters the optimal operation mode.

10. The boiler full-process collaborative intelligent control system according to claim 9, characterized in that: Dynamically correct the operating parameters according to the actual operating status of the boiler, which specifically includes the following steps: Detect whether there is any deviation based on the feedback status of the boiler; Dynamically correct boiler operating parameters based on deviations; After the boiler operating parameters are corrected, the actual operating status of the boiler is monitored in real time to verify the correction effect until the deviation is eliminated.

Images