Hotel intelligent management method and system based on big data association feedback

Abstract

The invention discloses a hotel intelligent management method and system based on big data association feedback, and relates to the technical field of hotel intelligent management, and the method comprises the steps: collecting hotel equipment operation data and environment association data to form an original data set, and carrying out the multi-level preprocessing to generate a standardized feature data set; constructing a ventilation equipment performance degradation model and an energy consumption prediction model based on the standardized feature data set, and outputting a comprehensive evaluation report; mining association rules in combination with the standardized feature data set, and extracting equipment fault association features and an energy consumption optimization rule set; then, generating a dynamic maintenance strategy and an energy consumption regulation and control scheme; and finally, performing closed-loop evaluation and model iteration, and updating management parameters and decision rules. The intelligent hotel management system comprises an acquisition module, a data processing module, an optimization module and a management module, all the modules cooperate with one another, intelligent hotel management is achieved, and the management efficiency and the energy consumption optimization level are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hotel intelligent management, and particularly relates to a hotel intelligent management method and system based on big data correlation feedback. BACKGROUND

[0002] With the development of hotel intelligence, the demand for equipment state monitoring, energy consumption optimization and dynamic management is increasingly urgent, and the existing management mode has been difficult to meet the efficient operation demand.

[0003] The existing hotel intelligent management method and system based on big data correlation feedback rely on experience judgment for equipment maintenance, have hysteresis, lack precise basis for energy consumption control, and are prone to resource waste. At the same time, the data of hotel equipment operation, environmental change and personnel activity are scattered, and it is difficult to realize effective correlation analysis, resulting in low management efficiency and fluctuation of service quality. Therefore, it is necessary to provide a hotel intelligent management method and system based on big data correlation feedback to solve the above problems. SUMMARY

[0004] To solve the above technical problems, a hotel intelligent management method and system based on big data correlation feedback are provided, which solves the problem of the existing hotel intelligent management method and system based on big data correlation feedback relying on experience judgment for equipment maintenance, having hysteresis, lacking precise basis for energy consumption control, and being prone to resource waste. At the same time, the data of hotel equipment operation, environmental change and personnel activity are scattered, and it is difficult to realize effective correlation analysis, resulting in low management efficiency and fluctuation of service quality.

[0005] To achieve the above purposes, the technical scheme adopted by the present application is as follows: A hotel intelligent management method based on big data correlation feedback, comprising: S1. Collecting hotel equipment operation data and environmental correlation data to form an original data set, wherein the equipment operation data covers real-time parameters of ventilation equipment and real-time parameters of electrical equipment, and the environmental correlation data includes indoor environmental indicators, outdoor environmental indicators and personnel activity data; S2. Multi-level preprocessing of the original data set to generate a standardized feature data set, wherein the preprocessing includes data cleaning, anomaly filtering and feature reconstruction; S3. Building a ventilation equipment performance degradation model and an energy consumption prediction model based on the standardized feature data set, and outputting a comprehensive evaluation report; S4. Correlating the standardized feature data set with the comprehensive evaluation report to mine correlation rules and extract equipment fault correlation features and energy consumption optimization rule sets; S5. Generating dynamic maintenance strategies and energy consumption control schemes based on the correlation rules and real-time monitoring data; S6. Conduct closed-loop evaluation and model iteration of dynamic maintenance strategies and energy consumption control schemes, and update management parameters and decision-making rules.

[0006] In an optional embodiment, step S1 specifically includes: S101. Wind speed is collected via a built-in sensor in the ventilation equipment. Wind pressure Motor speed and cumulative runtime After integration, the real-time parameters of the ventilation equipment are obtained. ; S102. Collecting three-phase current of electrical equipment using smart meters and current transformers. , , ,Voltage and active power After integration, the real-time parameters of the electrical equipment are obtained. ; S103. Deploy temperature and humidity sensors to obtain indoor temperature. ,humidity With outdoor temperature ,humidity After integration, indoor environmental indicators are obtained; S104. Indoor CO2 concentration is collected using an air quality sensor. PM2.5 value After integration, outdoor environmental indicators are obtained. Then, indoor and outdoor environmental indicators are combined to obtain environmental correlation data. ; S105. Obtain room occupancy status using the guest room control system. Frequency of personnel activities After integration, the personnel activity data is obtained. ; S106. Collect historical maintenance records of hotel equipment, including repair time. Replacement of parts and fault codes After integration, the hotel's historical equipment status data is obtained. ; S107. Real-time parameters of ventilation equipment Real-time parameters of electrical equipment Environmental related data Personnel activity data and hotel historical equipment status data Summarize to form the original dataset .

[0007] In an optional embodiment, step S2 specifically includes: S201. Obtain the cumulative running time corresponding to the original data set , and obtain the original data corresponding to each time point t; S202. Extract the original data of the same type 0 to t moment in the original data set to obtain the first to-be-processed data; S203. Obtain the data missing time node in the first to-be-processed data , and then use the mean filling method to process the missing value: , wherein is the data missing time node in the first to-be-processed data which needs to be filled, k is the number of adjacent valid data points corresponding to the data missing time node , and is the i-th adjacent valid data value corresponding to the data missing time node ; S204. Integrate the first to-be-processed data after filling the missing value to obtain the second to-be-processed data, and then eliminate the abnormal value in the second to-be-processed data based on the 3σ criterion: when , it is marked as abnormal, wherein is the i-th second to-be-processed data of the same type, μ is the data mean value of the second to-be-processed data of the same type, and σ is the standard deviation of the second to-be-processed data of the same type. S205. Fourier transform the current signal in the second to-be-processed data after eliminating the abnormal value: , calculate the harmonic component and the total harmonic distortion rate ; S206. Obtain the power signal in the second to-be-processed data, and extract the power signal transient feature by using wavelet packet decomposition: , wherein is the wavelet coefficient, j is the decomposition layer number, is the translation parameter, is the wavelet basis function; S207. Based on the real-time parameters of the ventilation equipment in the second to-be-processed data , obtain the fresh air volume and fan power of the ventilation equipment, so as to determine the energy efficiency ratio of the ventilation equipment: , wherein is the fresh air volume, is the fan power; S208. Normalize the environment-related data : , map to the [0, 1] interval, wherein is the normalized standard value of the environment-related data, is the maximum value in the environment-related data, the minimum value in the environmental related data; S209. Constructing the equipment operation state feature matrix wherein m is the number of time windows, n is the feature dimension, is the element in the bth row and cth column of matrix X; S210. Using principal component analysis to reduce the dimension of the equipment operation state feature matrix, obtaining the standardized feature matrix: X red =X·W, wherein W is the feature vector matrix, X red is the standardized feature matrix; S211. Based on the standardized feature matrix, constructing the standardized feature data set.

[0008] In an optional embodiment, step S3 specifically comprises: S301. Calculating the equipment health index based on the random forest algorithm: wherein is the weight of the ith decision tree, is the health value output by the ith tree, is the total number of decision trees; S302. Using the LSTM network to predict the energy consumption in the next 24 hours: wherein is the energy consumption prediction value, is the prediction time point, is the prediction step; S303. Establishing the performance degradation model of the ventilation equipment: wherein is the current energy efficiency coefficient of the ventilation equipment, is the degradation coefficient; S304. Calculating the remaining service life of the equipment wherein, is the time when the equipment is expected to fail, is the current time, is calculated by the Weibull distribution: wherein, is the cumulative distribution function, is the time of equipment operation, is the scale parameter, is the shape parameter; S305. Constructing the energy consumption baseline wherein is the personnel density, is the indoor temperature, is the room occupancy status; S306. Outputting the equipment health level: wherein corresponding health status; S307. Generating energy consumption deviation early warning threshold: wherein is the energy consumption deviation early warning threshold, is the historical deviation mean, is the historical deviation standard deviation; S308. Integrating model output to form a comprehensive evaluation report: wherein is the equipment health index, is the equipment remaining service life, is the energy consumption prediction value at the prediction time point , is the energy consumption baseline, is the energy consumption deviation early warning threshold.

[0009] In an optional embodiment, step S4 specifically comprises: S401. Based on the comprehensive evaluation report, obtaining all prediction time points from 0 to t whose difference between and is greater than , and marking them as failure nodes; S402. Based on the failure nodes, analyzing contemporaneous equipment parameter changes and extracting equipment failure associated features from the standardized feature data set; S403. Using the FP-growth algorithm to mine failure associated item sets from the equipment failure associated features corresponding to the failure nodes: wherein X is a failure precursor, Y is a failure type, and N is the total number of data samples; S404. Calculating rule confidence: and screening rules with confidence greater than the association rule threshold; S405. Establishing a ventilation equipment wind pressure and pipeline resistance relationship model: wherein , are fitting coefficients, is the baseline pressure difference, is the flow rate of the fluid in the pipeline, is the ventilation equipment wind pressure; S406. Analyzing the correlation between power factor and harmonic content of electrical equipment: wherein φ is the phase difference; S407. Generating equipment maintenance cycle rules: wherein is the safety factor, is the equipment remaining service life, is the equipment maintenance cycle, is the equipment health standard index; ​S408. Constructing the personnel density and equipment load correlation matrix: where a is the number of areas, b is the number of equipment types, represents the area , equipment type corresponding personnel density, represents the area , equipment type corresponding equipment power, represents the personnel density-equipment power correlation matrix. S409. Output the energy consumption optimization rule set through the ventilation equipment wind pressure and pipeline resistance relationship model in step S405, the correlation of the power factor and harmonic content of the electrical equipment analyzed in step S406, the equipment maintenance cycle rule in step S407, and the personnel density and equipment load correlation matrix. Each rule contains a trigger condition and a regulation parameter.

[0010] In an optional embodiment, step S5 specifically comprises: S501. For area f, calculate the maintenance priority based on the health index and RUL: , is the maintenance priority; S502. Develop a dynamic adjustment strategy for ventilation equipment: where , , are the baseline values corresponding to each data pair, respectively; S503. Design a failure emergency response process: when trigger the standby equipment switching, otherwise do not trigger, where is the preset health index threshold; S504. Develop energy consumption over-standard intervention measures: when start load reduction , where is the amount of load that needs to be reduced; S505. Generate a dynamic maintenance strategy, and then optimize the equipment operation mode combination: satisfying the total power constraint , where is the energy efficiency coefficient of the s-th optimized equipment in area f, is the operating power of the s-th optimized equipment in area f, is the total operating power of the equipment group in area f, is the total power constraint threshold, is the total number of optimized equipment in area f; S506. Output the energy consumption regulation scheme.

[0011] In an optional embodiment, step S6 specifically comprises: S601. After the dynamic maintenance strategy and the energy consumption regulation scheme, calculate the dynamic maintenance strategy execution effect index: wherein is the strategy execution comprehensive effect index, are weight coefficients, is the energy consumption change amount, is the health index change amount, is the remaining life change amount; S602. Analyze the energy consumption reduction rate: wherein is the energy consumption reduction rate, is the energy consumption before strategy execution, is the energy consumption after strategy execution; S603. Statistics failure occurrence rate change: wherein is the failure occurrence rate change rate, is the number of failures before strategy execution, is the number of failures after strategy execution; S604. Update model parameters using incremental learning: wherein is the updated model parameter, is the model parameter before updating, is the learning rate, is the gradient of the loss function, is the new training data; S605. Adjust the health index calculation weight: wherein is the adjusted health index weight, is the health index weight before adjustment, is the weight adjustment coefficient, is the effect index change rate; S606. Optimize the association rule threshold: wherein is the optimized association rule support threshold, is the association rule support threshold before optimization, is the threshold adjustment coefficient, is the failure occurrence rate change rate; S607. Generate model iteration report: wherein is the model iteration report, is the energy consumption reduction rate, failure change rate, is the updated model parameter, is the optimized association rule threshold; S608. Output the updated management parameter set: wherein is the updated management parameter set, is the model parameter, is the rule parameter, is the threshold parameter.

[0012] Further, a hotel intelligent management system based on big data correlation feedback is proposed, which is used to implement the management method of any one of the above, comprising: A collection module is configured to collect hotel equipment operation data and environment correlation data to form an original data set; A data processing module is configured to perform multi-level preprocessing on the original data set to generate a standardized feature data set; An optimization module is configured to construct a ventilation equipment performance degradation model and an energy consumption prediction model based on the standardized feature data set, and output a comprehensive evaluation report, and is further configured to combine the standardized feature data set to perform association rule mining on the comprehensive evaluation report, and extract equipment fault correlation features and energy consumption optimization rule sets; A management module is configured to combine the association rules and real-time monitoring data to generate dynamic maintenance strategies and energy consumption control schemes, and is further configured to perform closed-loop evaluation and model iteration on the dynamic maintenance strategies and energy consumption control schemes, and update the management parameters and decision rules.

[0013] In an optional embodiment, the optimization module comprises: A model construction unit is configured to construct a ventilation equipment performance degradation model and an energy consumption prediction model based on the standardized feature data set, and output a comprehensive evaluation report; A feature and rule mining unit is configured to combine the standardized feature data set to perform association rule mining on the comprehensive evaluation report, and extract equipment fault correlation features and energy consumption optimization rule sets.

[0014] In an optional embodiment, the management module comprises: A strategy execution unit is configured to combine the association rules and real-time monitoring data to generate dynamic maintenance strategies and energy consumption control schemes; An update unit is configured to perform closed-loop evaluation and model iteration on the dynamic maintenance strategies and energy consumption control schemes, and update the management parameters and decision rules.

[0015] Compared with the prior art, the present application has the following advantages: The hotel intelligent management method based on big data correlation feedback provided in the scheme realizes the standardization and precision of data by collecting hotel equipment operation data and environment correlation data and performing multi-level preprocessing, provides a high-quality data basis for subsequent model construction, and improves the reliability of analysis results; The hotel intelligent management method based on big data correlation feedback provided in the scheme realizes accurate grasping of the equipment state and energy consumption by constructing a ventilation equipment performance degradation model and an energy consumption prediction model and outputting a comprehensive evaluation report, can know the equipment health status and energy consumption trend in advance, and is convenient for early planning and management; The hotel intelligent management method based on big data correlation feedback provided in the scheme realizes the pertinence of fault early warning and energy consumption regulation by mining association rules to extract equipment fault correlation characteristics and an energy consumption optimization rule set, can reduce faults and reduce energy consumption cost; The hotel intelligent management method based on big data correlation feedback provided in the scheme realizes the dynamic and continuous optimization of management by generating dynamic maintenance strategies and energy consumption regulation schemes and performing closed-loop evaluation iteration, can adapt to hotel operation changes, and continuously improves management efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 The flowchart of the hotel intelligent management method based on big data correlation feedback provided in the present application is shown in the figure; Figure 2 The flowchart of the acquisition of the standardized feature data set in the present application is shown in the figure; Figure 3 The flowchart of the acquisition of the comprehensive evaluation report in the present application is shown in the figure; Figure 4 The system framework diagram of the hotel intelligent management system based on big data correlation feedback provided in the present application is shown in the figure. DETAILED DESCRIPTION

[0017] The following description is used to disclose the present application so that those skilled in the art can implement the present application. The preferred embodiments in the following description are only as examples, and other obvious modifications can be thought of by those skilled in the art.

[0018] REFERENCE Figure 1 - Figure 4 As shown in the figure, the hotel intelligent management method based on big data correlation feedback comprises: S1. Collecting hotel equipment operation data and environment correlation data to form an original data set, the equipment operation data covering real-time parameters of ventilation equipment and real-time parameters of electrical equipment, and the environment correlation data including indoor environment indexes, outdoor environment indexes and personnel activity data; S2. Perform multi-level preprocessing on the original dataset to generate a standardized feature dataset. The preprocessing includes data cleaning, anomaly filtering, and feature reconstruction. S3. Construct a performance degradation model and energy consumption prediction model for ventilation equipment based on a standardized feature dataset, and output a comprehensive evaluation report; S4. Combine the standardized feature dataset to perform association rule mining on the comprehensive evaluation report and extract equipment fault association features and energy consumption optimization rule set; S5. Combine association rules with real-time monitoring data to generate dynamic maintenance strategies and energy consumption control schemes; S6. Conduct closed-loop evaluation and model iteration of dynamic maintenance strategies and energy consumption control schemes, and update management parameters and decision-making rules.

[0019] Furthermore, step S1 specifically includes: S101. Wind speed is collected via a built-in sensor in the ventilation equipment. Wind pressure Motor speed and cumulative runtime After integration, the real-time parameters of the ventilation equipment are obtained. ; S102. Collecting three-phase current of electrical equipment using smart meters and current transformers. , , ,Voltage and active power After integration, the real-time parameters of the electrical equipment are obtained. ; S103. Deploy temperature and humidity sensors to obtain indoor temperature. ,humidity With outdoor temperature ,humidity After integration, indoor environmental indicators are obtained; S104. Indoor CO2 concentration is collected using an air quality sensor. PM2.5 value After integration, outdoor environmental indicators are obtained. Then, indoor and outdoor environmental indicators are combined to obtain environmental correlation data. ; S105. Obtain room occupancy status using the guest room control system. Frequency of personnel activities After integration, the personnel activity data is obtained. ; S106. Collect historical maintenance records of hotel equipment, including repair time. Replacement of parts and fault codes hotel historical equipment state data ; S107. aggregating the ventilation equipment real-time parameters , the electrical equipment real-time parameters , the environment-related data , the personnel activity data and the hotel historical equipment state data to form an original data set .

[0020] Specifically, in the data collection process, through ventilation equipment built-in sensors, smart meters, current transformers, temperature and humidity sensors, air quality sensors, guest room control systems and other equipment, ventilation equipment parameters, electrical equipment parameters, indoor and outdoor environmental indicators, personnel activity data and equipment historical maintenance records are collected respectively, and then these data are aggregated to form an original data set. Ventilation equipment parameters such as wind speed and wind pressure reflect the running state of the ventilation equipment; electrical parameters such as three-phase current and voltage reflect the energy consumption of electrical equipment. Environmental data such as indoor and outdoor temperature and humidity, CO2 concentration reflect the environmental quality; room occupancy status (such as whether there is a person), personnel activity frequency (such as the number of personnel entering and exiting) reflect the influence of personnel activity on hotel resource demand. Equipment historical maintenance records provide a reference for equipment failure prediction and maintenance.

[0021] The advantage is that: through multi-type sensors and system collection of multi-dimensional data, the comprehensive coverage of hotel operation related data is realized, the data island is broken, and rich and complete basic data are provided for subsequent analysis. The collected data are integrated into ventilation equipment parameters, environment-related data and other modules according to categories, realizing systematic classification of data, making data structure clear, and facilitating subsequent targeted processing and analysis. The non-real-time static data such as equipment historical maintenance records and personnel activity data are included, realizing the combination of real-time dynamic data and historical static data, enhancing the correlation and historical reference value of data, and improving the depth and accuracy of subsequent model analysis.

[0022] Further, step S2 specifically comprises: S201. obtaining the cumulative running time corresponding to the original data set , and obtaining the original data corresponding to each time point t; S202. extracting the original data of the same type from 0 to t moment in the original data set to obtain first to-be-processed data; S203. obtaining the data missing time nodes in the first to-be-processed data , and then using the mean filling method to process the missing values: , wherein, is the data missing time node in the first to-be-processed data Missing values need to be filled in, k is the data missing time node The number of corresponding adjacent valid data points, is the data missing time node The corresponding i-th adjacent valid data value; S204. The first to-be-processed data after filling in the missing values is integrated to obtain the second to-be-processed data, and then the abnormal values in the second to-be-processed data are removed based on the 3σ criterion: when is marked as abnormal, wherein, is the i-th second to-be-processed data of the same type, μ is the data mean of the second to-be-processed data of the same type, and σ is the standard deviation of the second to-be-processed data of the same type. S205. Fourier transform is performed on the current signal in the second to-be-processed data after removing the abnormal values: , the harmonic component and the total harmonic distortion rate are calculated. S206. The power signal in the second to-be-processed data is obtained, and the wavelet packet decomposition is used to extract the transient characteristics of the power signal: , wherein, is the wavelet coefficient, j is the decomposition level, is the translation parameter, is the wavelet basis function. S207. Based on the real-time parameters of the ventilation equipment in the second to-be-processed data , the fresh air volume and the fan power of the ventilation equipment are obtained, so as to determine the energy efficiency ratio of the ventilation equipment: , wherein is the fresh air volume, is the fan power. S208. The environment-related data is normalized: , which is mapped to the interval [0, 1], wherein, is the normalized standard value of the environment-related data, is the maximum value in the environment-related data, is the minimum value in the environment-related data. S209. The equipment operation state feature matrix is constructed , wherein m is the number of time windows, n is the feature dimension, is the element in the b-th row and c-th column position of the matrix X. S210. The principal component analysis is used to reduce the feature dimension of the equipment operation state feature matrix to obtain the standardized feature matrix: X red =X·W, wherein W is the feature vector matrix, and X red is the standardized feature matrix. S211. Construct a standardized feature dataset based on the standardized feature matrix.

[0023] Specifically, in steps S201-S202, the cumulative running time corresponding to the original dataset and the original data at each time point are first obtained, and then the data at the same type 0 to t time is extracted to obtain the first to-be-processed data. This step is a preliminary screening of the original data according to time and type, which clearly defines the object for subsequent processing and ensures the time continuity and type consistency of the data. In S203, missing time nodes in the first to-be-processed data are identified and missing values are processed using the mean filling method (taking the average of the adjacent k valid data to fill in the missing values). This can reduce the impact of data missing on subsequent analysis and ensure data integrity. In step S204, the filled data is integrated to obtain the second to-be-processed data, and abnormal values are removed based on the 3σ criterion (when the data deviates from the mean by more than 3 times the standard deviation, it is considered abnormal). This can filter extreme abnormal data and improve data reliability. In steps S205-S206, Fourier transform is performed on the current signal to obtain harmonic components and distortion rate, and wavelet packet decomposition is used to extract transient features from the power signal. This can mine the device running state information implied in the electrical signal and provide a basis for device fault diagnosis. In step S207, the energy efficiency ratio (the product of the fan power and the fresh air volume) is calculated according to the ventilation equipment parameters, which intuitively reflects the energy utilization efficiency of the ventilation equipment. In step S208, the environmental correlation data is normalized (mapped to the [0, 1] interval), which eliminates the dimensional differences of different environmental indicators and facilitates cross-indicator analysis. In steps S209-S210, the device running state feature matrix is constructed, and principal component analysis is used to reduce the dimension to obtain the standardized feature matrix. This can simplify the data dimension, retain key information, and improve the operation efficiency of subsequent models. In step S211, a standardized feature dataset is constructed based on the standardized feature matrix, which provides a high-quality, standardized data basis for subsequent model construction and rule mining.

[0024] The advantages are that by filling in missing values with the mean and removing abnormal values with the 3σ criterion, data cleaning and correction are achieved, ensuring data quality and providing reliable input for subsequent analysis. By performing Fourier transform, wavelet packet decomposition, and calculating energy efficiency ratio on electrical signals, feature extraction and transformation of data are achieved, deep features of data are mined, and the analysis value of data is enhanced. By normalizing, dimensionality reduction, and other processing to construct a standardized feature dataset, data standardization and simplification are achieved, which unifies the data format, reduces redundancy, and improves the efficiency and accuracy of subsequent model construction.

[0025] Further, step S3 specifically includes: S301. Calculate the device health index based on the random forest algorithm: wherein is the weight of the i-th decision tree, is the health value output by the i-th tree. total number of decision trees; Specifically, S302. Predicting future 24-hour energy consumption using an LSTM network: wherein is the energy consumption prediction value, is the prediction time point, is the prediction step, is the environmental parameter, the environmental parameter (or other external factors affecting energy consumption, combined with the hotel scene, such as indoor and outdoor temperature, humidity, personnel density, and other environmental data), =1, 2, …, 24; S303. Establishing a ventilation equipment performance degradation model: wherein is the current energy efficiency coefficient of the ventilation equipment, is the degradation coefficient; Specifically, is the current energy efficiency coefficient of the ventilation equipment, directly measured (physical sensor): For air conditioners: through “refrigerating capacity sensor + power meter”, calculate η(t)=refrigerating capacity(t) / input power(t) (i.e. energy efficiency ratio EER); for fans: through “air volume sensor + power meter”, calculate η(t)=air supply volume(t) / input power(t) (i.e. fan efficiency); is the initial efficiency of the equipment, which is the “baseline efficiency” when the equipment is brand new (or just after maintenance), which is the starting point of performance degradation, and its existence is factory calibration: the equipment manufacturer will provide “efficiency parameters of new equipment” (such as EER value on air conditioner nameplate), which is directly used as After maintenance calibration: after equipment overhaul / maintenance, re-measure efficiency (method same as direct measurement of ), the efficiency at this time is taken as the new (cover historical degradation, reset baseline). The degradation coefficient is fitted by historical data, collect the efficiency data and corresponding running time of the equipment throughout its life cycle, use exponential regression model for fitting, then solve the degradation coefficient by least square method.

[0026] S304. Calculate the remaining service life of the equipment wherein, is the time when the equipment is expected to fail, is the current time, is calculated by Weibull distribution: wherein, is the cumulative distribution function, is the time of equipment operation, is the scale parameter, is the shape parameter, i.e. the remaining useful life, in scenarios such as hotel equipment intelligent management (ventilation, electrical equipment, etc.), to represent the length of time that the equipment can still operate normally from the current time to the time when the failure (or performance degradation to the point of not being able to work normally) is expected to occur, representing the time when the equipment is expected to fail (break down), is the time point obtained by predicting the equipment failure time through a prediction model (such as Weibull distribution, machine learning model, etc.), used to determine when the equipment is likely to fail, so as to plan maintenance, replacement, etc. in advance, represents the current time, which is the reference time point for calculating the remaining useful life, reflecting the starting point of the remaining time calculation from the current time to the time when the equipment is expected to fail; Specifically, represents the cumulative distribution function, which is used to describe the probability of equipment failure (breakdown) before time t. That is, the cumulative probability of equipment failure at time t is F(t), for example, F(100)=0.2, which means that the probability of equipment failure within 100 time units (hours, days, etc. depending on the scene) is 20%. is the scale parameter, also known as characteristic life. It determines the "stretching" degree of the Weibull distribution, affects the shape of the distribution curve, and reflects the rate of change of the equipment failure probability over time. Under different equipment or different operating conditions, the value is different, the greater the value, the more likely the equipment will fail after a longer period of time. is the shape parameter, also known as the Weibull slope. It plays a key role in the shape of the distribution curve, determining the trend of the equipment failure probability over time: When β=1, the Weibull distribution degenerates into an exponential distribution, meaning that the equipment failure probability is constant, with no significant wear or aging acceleration (similar to random failure); When β>1, as time t increases, the failure probability gradually increases, representing the cumulative effects of wear and aging of the equipment, and the longer the time, the more likely the failure (such as fatigue failure of mechanical parts); When β<1, the failure probability decreases with time, which may correspond to the early failure stage of the equipment (early failure period of bathtub curve), with a high initial failure probability, and the failure probability decreases after the equipment is "broken in".

[0027] S305. Constructing energy consumption baseline wherein is the personnel density, is the indoor temperature, is the room occupancy status; S306. Output equipment health level: wherein corresponds to the health status; S307. Generating an energy consumption deviation early warning threshold: wherein is the energy consumption deviation early warning threshold, is the historical deviation mean, is the historical deviation standard deviation; Specifically, the deviation early warning threshold is when the actual energy consumption (or other monitored indicators such as power, efficiency, etc.) of the device deviates from the benchmark value beyond the threshold, the system will trigger an early warning, suggesting that there may be abnormalities (such as abnormal fluctuations in energy consumption, precursors of device failure, etc.), for abnormal detection and early warning decision-making in device management. The historical deviation mean, i.e., the average value of the deviation between the actual value and the benchmark value (such as the energy consumption benchmark, normal operation parameter benchmark, etc.) of the device historical operation data, reflects the average level of the deviation of the device in the normal operation state, and is the basic reference quantity for constructing the early warning threshold. The historical deviation standard deviation is used to measure the dispersion degree and fluctuation range of the historical operation deviation of the device, and embodies the dispersion of the deviation data. The larger the standard deviation, the more intense the fluctuation of the historical deviation.

[0028] S308. Integrating model output to form a comprehensive evaluation report: wherein is the device health index, is the remaining service life of the device, is the energy consumption prediction value at the prediction time point represents the prediction result of the device energy consumption at time t, which is used to know the energy consumption of the device at a certain time in the future, to assist energy consumption regulation, cost estimation, etc., is the energy consumption benchmark, which refers to the energy consumption reference value of the device at time t under normal operating conditions, which can be obtained by historical normal data, standard working condition calculation, etc., and is used to compare with the actual energy consumption or predicted energy consumption to judge whether the energy consumption is abnormal, is the energy consumption deviation early warning threshold, which is used to define the reasonable range of energy consumption (or other indicators) deviation. When the deviation between the actual energy consumption and the benchmark energy consumption exceeds the threshold, it is determined to be abnormal, triggering an early warning to assist in timely discovery of device failure, energy efficiency abnormalities, etc. Specifically, the comprehensive evaluation report (or result set) is a summary of multi-dimensional information such as equipment operating status and performance prediction, which is used to comprehensively present key state indicators of the equipment at present and in the future period of time, and assist in equipment management decision-making (such as maintenance plan formulation, energy consumption optimization, etc.). The equipment health index is a quantitative index obtained through data analysis and model calculation, which is used to evaluate the current health status of the equipment. The higher the value (or according to the set rules), the healthier the equipment, which can reflect the equipment wear, aging, fault hidden danger, etc. For example, a health index of 0.8 may represent a good equipment state, and a health index of 0.3 may indicate that there are more fault hidden dangers. The remaining useful life represents the length of time that the equipment can continue to work from the current time to the time when it is predicted to be unable to operate normally (failure, performance not meeting requirements, etc.), which is used to predict when the equipment may need maintenance or replacement, and to plan maintenance resources in advance.

[0029] The advantages are that the health index and predicted energy consumption are calculated by the random forest algorithm and the LSTM network respectively, realizing accurate quantification of equipment state and energy consumption, and providing scientific and quantifiable basis for management decision-making. By building a performance degradation model, calculating the remaining life and setting the health level, the state tracking of the equipment throughout its life cycle is realized, potential problems of the equipment can be found in advance, and the risk of sudden failure is reduced. By integrating multi-dimensional indicators to form a comprehensive evaluation report, information aggregation of equipment management and energy consumption control is realized, decision-making efficiency is improved, and it is convenient to make overall optimization strategy.

[0030] Further, step S4 specifically includes: S401. Based on the comprehensive evaluation report, all The difference between the predicted time point of greater than is recorded as a fault node; S402. Based on the fault node, the change of the device parameter in the same period is analyzed, and the device fault associated feature is extracted from the standardized feature data set; S403. The device fault associated feature corresponding to the fault node is mined for fault associated item set using the FP-growth algorithm: , wherein X is a fault precursor, Y is a fault type, and N is the total number of data samples; S404. Calculate the rule confidence: , and filter the rules with confidence greater than the association rule threshold; S405. Establish a model of the relationship between the air pressure of the ventilation equipment and the pipe resistance: , wherein , are fitting coefficients, is the reference pressure difference, is the flow rate of the fluid in the pipe, For the wind pressure of the ventilation equipment; S406. Analyzing the correlation between the power factor and harmonic content of the electrical equipment: wherein φ is the phase difference; S407. Generating the equipment maintenance cycle rule: wherein is the safety factor, is the remaining service life of the equipment, is the equipment maintenance cycle, is the equipment health standard index; S408. Building the personnel density and equipment load correlation matrix: wherein a is the number of areas, b is the type of equipment, indicating the dimension and structure of the matrix, and explaining is an a-row (number of areas) and b-column (number of equipment types) matrix, each element corresponding to a "person-power" correlation value of an area-equipment type, represents the area , equipment type corresponding to the personnel density (such as the occupancy density when the air conditioner in the guest room area is turned on), is the area identifier (such as the "guest room area, lobby, restaurant" and other different functional areas of the hotel, , a is the total number of areas), represents the area , equipment type (equipment type identifier (such as "air conditioner, lighting, ventilation machine" and other different equipment categories) corresponding to the equipment power, represents the personnel density-equipment power correlation matrix, which is used to integrate the correlation data of "personnel distribution" and "equipment energy consumption / power", and is the basic data structure for subsequent analysis of "how personnel activities affect equipment energy consumption"; S409. Through the wind pressure and pipe resistance relationship model of the ventilation equipment in step S405, the correlation between the power factor and harmonic content of the electrical equipment in step S406, the equipment maintenance cycle rule in step S407, and the personnel density and equipment load correlation matrix, output the energy consumption optimization rule set , each rule contains a trigger condition and a control parameter.

[0031] Specifically, in steps S401-S402, based on the comprehensive evaluation report, the time point with the energy consumption prediction value and the baseline difference exceeding the early warning threshold is screened out as the fault node, and then the fault correlation features are extracted from the standardized feature data set by analyzing the changes of the device parameters at the same period. The device parameter features related to the fault can be accurately located, providing clues for fault diagnosis. In steps S403-S404, the FP-growth algorithm is used to mine the fault correlation item set, and effective rules are screened through support (the probability of the simultaneous occurrence of the fault precursor and the type) and confidence (the probability of the occurrence of the fault when the precursor appears). The potential law of fault occurrence can be identified, and the accuracy of fault warning can be improved. The relationship model between the air pressure of the ventilation equipment and the pipeline resistance in step S405 reflects the quantitative relationship between the air pressure, the flow rate and the pipeline resistance, providing a theoretical basis for the regulation of the ventilation equipment. In step S406, the correlation between the power factor and the harmonic content of the electrical equipment is analyzed, the influence of harmonics on the energy efficiency of the equipment is revealed, and the electrical efficiency is improved.

[0032] The advantages are that by mining the fault correlation rules, the accurate matching of the fault precursor and the type is realized, the potential fault can be warned in advance, and the loss caused by the sudden fault is reduced. By establishing the correlation model between the device parameters and the energy consumption, the quantification of energy consumption optimization is realized, the energy consumption regulation is more targeted and scientific, and the energy waste is reduced. By integrating the multi-dimensional correlation rules to form an optimized rule set, the coordination of device maintenance and energy consumption control is realized, and the systematicness and efficiency of hotel operation management are improved.

[0033] Further, step S5 specifically includes: S501. For the region f, calculate the maintenance priority based on the health index and RUL: , for the maintenance priority; S502. Develop a dynamic adjustment strategy for the ventilation equipment: wherein , , are the baseline values corresponding to the respective data; S503. Design a fault emergency response process: when trigger the standby equipment switching, otherwise do not trigger, wherein is a preset health index threshold; S504. Develop energy consumption intervention measures: when start load reduction , wherein is the load amount that needs to be reduced; S505. Generate a dynamic maintenance strategy W = {ID (device ID), P main , component list, estimated man-hours, risk level}, and then optimize the device operation mode combination: , meet the total power constraint , wherein is the energy efficiency coefficient of the s-th to-be-optimized device in the region f, is the operating power of the s-th to-be-optimized device in the region f, is the total operating power of the device group in the region f, is the total power constraint threshold, is the total number of to-be-optimized devices in the region f; S506. Output the energy consumption regulation scheme L={t exec , execution parameters, expected effects, monitoring indicators}.

[0034] Specifically, in step S501, the priority quantifies the urgency of the device maintenance, the lower the health index and the shorter the remaining life, the higher the priority, which provides a basis for resource allocation. In step S502, the strategy dynamically adjusts the wind speed according to the real-time personnel density p(t) and CO concentration C CO2 (t), ensuring indoor air quality while reducing energy consumption. For step S503, a fault emergency response process is designed, which triggers a backup device switching when the device health index H is lower than the preset threshold H th , avoiding the impact of device failure on normal operation and improving system reliability. When the real-time energy consumption E(t) exceeds 1.1 times of the baseline E base (t), the load reduction amount is calculated by to timely curb energy consumption. In step S505, a dynamic maintenance strategy containing device ID, maintenance priority, component list, etc. is generated, and the device operation mode combination is optimized, maximizing the sum of device energy efficiency coefficient and power product in the region under the constraint that the total power does not exceed the threshold P limit , achieving energy efficiency optimization. In step S506, the energy consumption regulation scheme containing execution time, parameters, expected effects, etc. is output to provide clear guidance for implementation.

[0035] Further, step S6 specifically includes: S601. After the dynamic maintenance strategy and energy consumption regulation scheme, calculate the dynamic maintenance strategy execution effect index: , wherein is the strategy execution comprehensive effect index, are weight coefficients, is the energy consumption change amount (negative for reduction, positive for increase), is the health index change amount (positive for improvement), is the remaining life change amount (positive for extension); S602. Analyze the energy consumption reduction rate: (positive for extension) , wherein energy consumption before policy execution, energy consumption before policy execution, energy consumption after policy execution; S603. Calculate the failure rate change: wherein failure rate change rate, number of failures before policy execution, number of failures after policy execution; S604. Update model parameters using incremental learning: wherein updated model parameters, model parameters before update, learning rate, gradient of loss function, new training data; S605. Adjust the health index calculation weight: wherein adjusted health index weight, health index weight before adjustment, weight adjustment coefficient, effectiveness index change rate; S606. Optimize the association rule threshold: wherein optimized association rule support threshold, association rule support threshold before optimization, threshold adjustment coefficient, failure rate change rate; S607. Generate model iteration report: wherein model iteration report, energy consumption reduction rate, failure change rate, updated model parameters, optimized association rule threshold; S608. Output updated management parameter set: wherein updated management parameter set, model parameters, rule parameters, threshold parameters.

[0036] Specifically, the comprehensive effect index is used to evaluate the overall benefit of the strategy execution. Steps S602-S603 are used to quantify the specific effects of the strategy in terms of energy saving and fault control. Step S604 adjusts the model parameters by adding new training data Dnew (data collected and processed by the acquisition module or data of the same type from the network) to adapt the model to new data changes and maintain analysis accuracy. The strategy effect is evaluated by multiple indicators, which realizes the quantitative feedback of management effect, clearly defines the pros and cons of the strategy, and provides a targeted direction for iteration. Incremental learning is used to update the model and parameters, which realizes the self-evolution of the system, so that the model continuously adapts to changes in hotel operation and maintains long-term effectiveness. The updated management parameter set is output, which realizes the closed-loop optimization of management decision-making, forms a virtuous cycle of "execution-evaluation-optimization", and continuously improves management accuracy and efficiency.

[0037] Further, a hotel intelligent management system based on big data correlation feedback is proposed, which is used to implement the management method of any one of the above, comprising: An acquisition module is used to collect hotel equipment operation data and environment correlation data to form an original data set. A data processing module is used to perform multi-level preprocessing on the original data set to generate a standardized feature data set. An optimization module is used to construct a ventilation equipment performance degradation model and an energy consumption prediction model based on the standardized feature data set, and output a comprehensive evaluation report. It is also used to combine the standardized feature data set to mine the correlation rules of the comprehensive evaluation report, and extract the equipment fault correlation features and energy consumption optimization rule set. A management module is used to combine the correlation rules and real-time monitoring data to generate dynamic maintenance strategies and energy consumption control schemes, and is also used for closed-loop evaluation and model iteration of the dynamic maintenance strategies and energy consumption control schemes, updating management parameters and decision rules.

[0038] Further, the optimization module comprises: A model construction unit is used to construct a ventilation equipment performance degradation model and an energy consumption prediction model based on the standardized feature data set, and output a comprehensive evaluation report. A feature and rule mining unit is used to combine the standardized feature data set to mine the correlation rules of the comprehensive evaluation report, and extract the equipment fault correlation features and energy consumption optimization rule set.

[0039] Further, the management module comprises: A strategy execution unit is used to combine the correlation rules and real-time monitoring data to generate dynamic maintenance strategies and energy consumption control schemes. An updating unit is configured to perform closed-loop evaluation and model iteration on the dynamic maintenance strategy and the energy consumption regulation scheme, and update the management parameters and the decision rules.

[0040] The above shows and describes the basic principles, main features and advantages of the present application. Those skilled in the art should understand that the present application is not limited to the above-mentioned embodiments, and the above-mentioned embodiments and descriptions in the specification are only the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A hotel intelligent management method based on big data correlation feedback, characterized in that, include: S1. Collect hotel equipment operation data and environmental correlation data to form a raw dataset. The equipment operation data includes real-time parameters of ventilation equipment and real-time parameters of electrical equipment. The environmental correlation data includes indoor environmental indicators, outdoor environmental indicators and personnel activity data. S2. Perform multi-level preprocessing on the original dataset to generate a standardized feature dataset. The preprocessing includes data cleaning, anomaly filtering, and feature reconstruction. S3. Construct a performance degradation model and energy consumption prediction model for ventilation equipment based on a standardized feature dataset, and output a comprehensive evaluation report; S4. Combine the standardized feature dataset to perform association rule mining on the comprehensive evaluation report and extract equipment fault association features and energy consumption optimization rule set; S5. Combine association rules with real-time monitoring data to generate dynamic maintenance strategies and energy consumption control schemes; S6. Conduct closed-loop evaluation and model iteration of dynamic maintenance strategies and energy consumption control schemes, and update management parameters and decision rules.

2. The hotel intelligent management method based on big data correlation feedback according to claim 1, characterized in that, Step S1 specifically includes: S101. Wind speed is collected via a built-in sensor in the ventilation equipment. Wind pressure Motor speed and cumulative runtime After integration, the real-time parameters of the ventilation equipment are obtained. ; S102. Using smart meters and current transformers to collect the three-phase current of electrical equipment. ,Voltage and active power After integration, the real-time parameters of the electrical equipment are obtained. ; S103. Deploy temperature and humidity sensors to obtain indoor temperature. ,humidity With outdoor temperature ,humidity After integration, indoor environmental indicators are obtained; S104. Indoor CO2 concentration is collected using an air quality sensor. value After integration, outdoor environmental indicators are obtained. Then, indoor and outdoor environmental indicators are combined to obtain environmental correlation data. ; S105. Obtain room occupancy status using the guest room control system. Frequency of personnel activities After integration, the personnel activity data is obtained. ; S106. Collect historical maintenance records of hotel equipment, including repair time. Replacement of parts and fault codes After integration, the hotel's historical equipment status data is obtained. ; S107. Real-time parameters of ventilation equipment Real-time parameters of electrical equipment Environmental related data Personnel activity data Hotel historical equipment status data Summarize to form the original dataset .

3. The hotel intelligent management method based on big data correlation feedback according to claim 1, characterized in that, Step S2 specifically includes: S201. Obtain the cumulative runtime corresponding to the original dataset. And obtain the raw data corresponding to each time point t; S202. Extract the original data of the same type from time 0 to time t from the original dataset to obtain the first data to be processed; S203. Obtain the time nodes when data is missing from the first set of data to be processed. Then, the mean imputation method was used to handle missing values: ,in, The time point where data is missing in the first set of data to be processed. The missing values ​​to be filled, where k is the time point where the data is missing. The corresponding number of adjacent valid data points For data missing time points The corresponding i-th adjacent valid data value; S204. Integrate the first set of data to be processed after imputing missing values ​​to obtain the second set of data to be processed. Then, remove outliers from the second set of data to be processed based on the 3σ criterion: when When σ is marked as an anomaly, Let μ be the i-th second data to be processed of the same type, μ be the mean of the second data to be processed of the same type, and σ be the standard deviation of the second data to be processed of the same type. S205. Perform a Fourier transform on the current signal in the second set of data to be processed after removing outliers: Calculate the harmonic components of each order. and total harmonic distortion ; S206. Obtain the power signal from the second set of data to be processed, and extract the transient features of the power signal using wavelet packet decomposition: ,in, Here, represents the wavelet coefficients, j represents the decomposition level, and y represents the translation parameter. These are wavelet basis functions; S207. Based on the real-time parameters of the ventilation equipment in the second data to be processed. To determine the energy efficiency ratio of ventilation equipment, the fresh air volume and fan power are obtained. ,in For fresh air volume, This refers to the fan power. S208. Environmental Related Data Normalization is performed: Mapped to the interval [0,1], where, These are the standard values ​​after normalization of environmental correlation data. The maximum value in the environmental correlation data. This is the minimum value in the environmental data. S209. Constructing the Equipment Operating Status Feature Matrix Where m is the number of time windows and n is the feature dimension. Let X be the element at the b-th row and c-th column of matrix X; S210. Principal component analysis is used to reduce the dimensionality of the equipment operating state feature matrix, resulting in a standardized feature matrix: X red =X·W, where W is the eigenvector matrix, and X... red The standardized feature matrix; S211. Construct a standardized feature dataset based on the standardized feature matrix.

4. The hotel intelligent management method based on big data correlation feedback according to claim 1, characterized in that, Step S3 specifically includes: S301. Calculate device health index based on random forest algorithm: ,in Let be the weight of the i-th decision tree. Let be the health value output by the i-th tree, and s be the total number of decision trees; S302. Uses an LSTM network to predict energy consumption for the next 24 hours: ,in This is the predicted energy consumption value. To predict the time point, To predict the step size; S303. Establish a performance degradation model for ventilation equipment: ,in The current energy efficiency coefficient of the ventilation equipment. The degradation coefficient; S304. Calculate the remaining useful life of the equipment. ,in, The expected time of equipment failure. For the current moment, Calculated using the Weibull distribution: ,in, The cumulative distribution function is... For the equipment running time, For scale parameters, For shape parameters; S305. Establishing an energy consumption baseline ,in For personnel density, Indoor temperature, The room is currently occupied. S306. Output device health level: ,in Corresponding health status; S307. Generate energy consumption deviation early warning threshold: ,in The energy consumption deviation warning threshold, The historical mean deviation The standard deviation is the historical bias. S308. The integrated model output forms a comprehensive evaluation report: Where H represents the equipment health index, and RUL represents the remaining service life of the equipment. This is the predicted energy consumption value at time point t. As a baseline for energy consumption, This is the early warning threshold for energy consumption deviation.

5. The hotel intelligent management method based on big data correlation feedback according to claim 1, characterized in that, Step S4 specifically includes: S401. Based on the comprehensive evaluation report, obtain all data within the range 0 to t. and The difference is greater than The predicted time point is denoted as the fault node; S402. Based on the correlation analysis of fault nodes, analyze the changes in equipment parameters during the same period, and extract equipment fault correlation features from the standardized feature dataset; S403. Use the FP-Growth algorithm to mine fault association itemsets from the equipment fault association features corresponding to the faulty nodes: Where X represents the precursor to a fault, Y represents the fault type, and N represents the total number of data samples; S404. Calculate the confidence level of the rule: Filter rules with a confidence level greater than the association rule threshold; S405. Establish a model for the relationship between ventilation equipment air pressure and duct resistance: ,in These are the fitting coefficients. As the reference pressure difference, The velocity of the fluid inside the pipe. For ventilation equipment air pressure; S406. Analyze the relationship between the power factor and harmonic content of electrical equipment: , where ϕ is the phase difference; S407. Generating Equipment Maintenance Cycle Rules: ,in For safety, RUL represents the remaining service life of the equipment. For equipment maintenance cycle, Equipment health standard index; S408. Construct a correlation matrix between personnel density and equipment load: Where 'a' represents the number of regions and 'b' represents the equipment type. This indicates the personnel density corresponding to region f and equipment type g. This indicates the power of the equipment corresponding to region f and equipment type g. This represents the personnel density-equipment power correlation matrix; S409. Using the ventilation equipment air pressure and duct resistance relationship model in step S405, the analysis of the correlation between power factor and harmonic content of electrical equipment in step S406, the equipment maintenance cycle rules in step S407, and the correlation matrix between personnel density and equipment load, output the energy consumption optimization rule set. Each rule includes triggering conditions and control parameters.

6. The hotel intelligent management method based on big data correlation feedback according to claim 1, characterized in that, Step S5 specifically includes: S501. For region f, calculate maintenance priority based on health index and RUL: , To maintain priority; S502. Develop dynamic adjustment strategies for ventilation equipment: ,in Each of these represents a baseline value corresponding to its respective data. S503. Design Fault Emergency Response Procedure: When If a switchover to a backup device is triggered, it will occur; otherwise, it will not. The preset health index threshold; S504. Develop intervention measures for excessive energy consumption: When At that time, load reduction is initiated. ,in, The amount of load that needs to be reduced; S505. Generate dynamic maintenance strategies, and then optimize the combination of equipment operating modes: Satisfying the total power constraint ,in Let be the energy efficiency coefficient of the s-th device to be optimized within region f. Let be the operating power of the s-th device to be optimized within region f. The total operating power of the equipment group within region f. fm is the total power constraint threshold, and fm is the total number of devices to be optimized in region f. S506. Output energy consumption control scheme.

7. The hotel intelligent management method based on big data correlation feedback according to claim 1, characterized in that, Step S6 specifically includes: S601. After implementing the dynamic maintenance strategy and energy consumption control scheme, calculate the performance indicators of the dynamic maintenance strategy: Where Effect is the overall performance indicator of the strategy execution. All are weighting coefficients. This represents the change in energy consumption. This represents the change in the health index. This represents the change in remaining lifetime. S602. Analysis of energy consumption reduction rate: ,in For energy consumption reduction rate, Energy consumption before strategy execution Energy consumption after strategy execution; S603. Statistical analysis of changes in failure rate: ,in The rate of change of failure rate The number of failures before the strategy is executed. This represents the number of failures after the strategy was executed. S604. Incremental learning is used to update model parameters: ,in For the updated model parameters, These are the model parameters before the update. For learning rate, The gradient of the loss function. To add new training data; S605. Adjust the weighting of the health index calculation: ,in For the adjusted health index weights, The original weights of the health index. This is the weighting adjustment factor. The rate of change of the performance indicator; S606. Optimize the threshold for association rules: ,in The optimized support threshold for association rules. The threshold for the support of association rules before optimization. This is the threshold adjustment coefficient. This represents the rate of change in the failure rate. S607. Generate model iteration report: ,in For model iteration reports, For energy consumption reduction rate and failure rate, For the updated model parameters, The optimized threshold for association rules; S608. Output the updated management parameter set: ,in For the updated management parameter set, For model parameters, For rule parameters, This is the threshold parameter.

8. A hotel intelligent management system based on big data correlation feedback, used to implement the management method as described in any one of claims 1-7, characterized in that, include: The data acquisition module is used to collect hotel equipment operation data and environmental correlation data to form a raw dataset; The data processing module is used to perform multi-level preprocessing on the original dataset to generate a standardized feature dataset; The optimization module is used to construct a performance degradation model and an energy consumption prediction model for ventilation equipment based on a standardized feature dataset, output a comprehensive evaluation report, and also to perform association rule mining on the comprehensive evaluation report in conjunction with the standardized feature dataset to extract equipment fault association features and an energy consumption optimization rule set. The management module is used to combine association rules and real-time monitoring data to generate dynamic maintenance strategies and energy consumption control schemes. It is also used to perform closed-loop evaluation and model iteration of dynamic maintenance strategies and energy consumption control schemes, and update management parameters and decision rules.

9. A hotel intelligent management system based on big data correlation feedback as described in claim 8, characterized in that, The optimization module includes: The model building unit is used to build a performance degradation model and an energy consumption prediction model for ventilation equipment based on a standardized feature dataset, and output a comprehensive evaluation report. The feature and rule mining unit is used to combine a standardized feature dataset to perform association rule mining on the comprehensive evaluation report, and extract equipment fault association features and energy consumption optimization rule sets.

10. A hotel intelligent management system based on big data correlation feedback as described in claim 8, characterized in that, The management module includes: A strategy execution unit is used to combine association rules and real-time monitoring data to generate dynamic maintenance strategies and energy consumption control schemes. The update unit is used to perform closed-loop evaluation and model iteration of dynamic maintenance strategies and energy consumption control schemes, and update management parameters and decision rules.

Images