Smart park big data comprehensive management and control system and method

Abstract

The invention discloses a smart park big data comprehensive management and control system, and the system comprises a data module which obtains the energy consumption data of a park and the external environment data of the park in real time, and carries out the processing of the energy consumption data and the environment data; and the energy management module takes the processed energy consumption data and environmental data as the input and output of a prediction model to obtain an energy optimization result, and sends the energy optimization result to the park service module and the multi-dimensional display module, and the multi-dimensional display module is used for displaying the energy optimization result in multiple dimensions. By acquiring the energy consumption data and the external environment data of the park in real time, the park can predict the future energy consumption condition based on the actual data and output the energy optimization result, so that the park can dynamically adjust the energy use strategy, the energy waste is reduced, and the energy utilization efficiency is improved. Meanwhile, personalized energy-saving schemes can be formulated according to behavior habits and requirements of different users, and the energy-saving awareness of the users can be improved.

Description

Technical Field

[0001] The present invention relates to the field of energy management and control technology, and specifically to a smart park big data comprehensive management and control system and method. Background Art

[0002] With the rapid economic development and the country's emphasis on energy conservation and emission reduction, industrial parks are undergoing a process of transformation from traditional development models to sustainable development driven by scientific and technological innovation. In this context, smart energy has become a key factor. It uses supercomputers and cloud computing technology to integrate the Internet of Things, forming new models and services with the characteristics of intelligent equipment, multi-functional collaboration, and information transparency, providing industrial parks with more sophisticated and intelligent energy management methods. In order to meet the new needs of park development, the smart energy management and control system can be used to ensure the park's energy security, centralized information management, intelligent detection network architecture and other services, helping the park to truly achieve "intelligence".

[0003] The Chinese invention patent with publication number CN116596190A discloses a smart park energy comprehensive management and control platform based on data analysis, including a collection unit, which is used to collect smart park energy data; a processing unit, which is communicated with the collection unit and the database. The processing unit integrates meteorological and energy consumption data on the basis of energy data collection, and uses big data and large sample analysis methods to intelligently classify, store, and process the acquired data and store them in the database; a monitoring unit, which is used to monitor the network status of each distributed energy station; the beneficial effects of the invention are: identifying equipment anomalies and locating abnormal equipment through the identification and positioning unit; through smart energy management, analyzing the park energy, abnormal conditions and warning conditions, providing scientific decision-making support, reducing the park operation cost, reducing costs and increasing efficiency, and ensuring the intelligent operation and management of the park.

[0004] In the process of management and control, the intelligent park energy comprehensive management and control platform of the above and similar technologies generally adopts the coordinated optimization control of the distribution network, distributed power supply, and diversified loads to improve. However, in the actual improvement process, park users often do not know how to save energy through personal behavior, which leads to the park's energy-saving behavior not starting from individuals. At the same time, the park cannot effectively combine external electricity prices to improve the park's energy input cost from the cost point of view and improve the park's comprehensive management capabilities, which leads to high energy input costs in the park and increases the economic pressure on the park. Summary of the invention

[0005] The purpose of the present invention is to provide a smart park big data comprehensive management and control system and method to solve the problems raised in the above background technology.

[0006] To achieve the above purpose, the present invention provides the following technical solutions: a smart park big data comprehensive management and control system, including:

[0007] A data module, which acquires the energy consumption data of the park and the external environment data of the park in real time, and processes the energy consumption data and the environment data, wherein the energy consumption data includes the personal energy consumption data of the user;

[0008] The energy management module uses the processed energy consumption data and environmental data as the input of the prediction model, and outputs the energy optimization results, which include energy scheduling solutions, personalized energy-saving solutions for parks, personalized energy-saving solutions for users, and working parameters of heating or cooling systems. At the same time, the energy optimization results are sent to the park service module and the multi-dimensional display module. The multi-dimensional display module is used to display the energy optimization results in multiple dimensions. The park service module includes:

[0009] A linkage unit, which determines the intelligent linkage mode between different devices in the park according to the energy scheduling plan and the park's personalized energy-saving plan;

[0010] Participating units formulate energy-saving incentive activities according to the user's personalized energy-saving plan, and the energy-saving incentive activities include energy-saving competitions, reward points, energy-saving challenges, energy-saving funds and energy-saving certification;

[0011] The adjusting unit adjusts the starting temperature threshold of the heating or cooling system according to the working parameters of the heating or cooling system.

[0012] Furthermore, it also includes an energy storage module electrically connected to the park equipment and the external power supply, which stores electric energy by identifying the power supply time period, specifically including:

[0013] M1: Extract characteristic values ​​from historical power consumption data and electricity price data, the characteristic values ​​include average electricity price, electricity price fluctuation range, electricity price standard deviation and electricity price change rate;

[0014] M2: Determine the power supply valley time period according to the characteristic value, specifically:

[0015] T low ={t|P(t)≤α·(μ+β·σ)}

[0016] Where: T low is the power supply off-peak period, t is the power supply time, P(t) is the electricity price corresponding to the power supply time t, α is a constant, μ is the average electricity price, β is a constant, and σ is the standard deviation of the electricity price;

[0017] M3: Determine whether the current time period is in the power supply valley time period. When it is in the power supply valley time period, electrically connect the energy storage battery in the energy storage module to the external power supply. When it is not in the power supply valley time period, electrically connect the energy storage battery in the energy storage module to the park equipment / facility.

[0018] Furthermore, the energy consumption data and environmental data are processed by methods including data cleaning and standardization.

[0019] Furthermore, the energy optimization result is obtained by:

[0020] NA1: construct feature engineering through the processed energy consumption data and environmental data, and the feature engineering includes time window features, seasonal features and weather features;

[0021] NA2: According to the feature engineering, multiple base models are constructed and a meta-feature matrix is ​​obtained. The multiple base models include a random forest model, a support vector machine model, a gradient boosting tree model and a long short-term memory network model;

[0022] NA3: construct at least one meta-model, and use the meta-feature matrix as an input of the meta-model to train the meta-model, wherein the meta-model includes a linear regression model, a logistic regression model, and a neural network model;

[0023] NA4: The energy consumption data and environmental data acquired and processed in real time are used as the input of the base model to obtain a real-time meta-feature matrix. Meanwhile, the real-time meta-feature matrix is ​​used as the input of the trained meta-model to output the corresponding energy optimization result.

[0024] Furthermore, obtaining the meta-feature matrix specifically includes:

[0025] NA2.1: Based on the feature engineering, multiple base models are constructed, and each base model is trained by a cross-validation method;

[0026] NA2.2: The prediction results corresponding to the base model training are used as meta-features, and the meta-features corresponding to each base model are combined to form a meta-feature matrix.

[0027] Furthermore, the method for obtaining the energy optimization result also includes:

[0028] NB1: construct feature engineering through the processed energy consumption data and environmental data, and the feature engineering includes time window features, seasonal features and weather features;

[0029] NB2: construct multiple base models, divide the feature engineering into a training set and a validation set, train each base model through the training set, and validate each trained base model through the validation set, wherein the multiple base models include a random forest model, a support vector machine model, a gradient boosting tree model, and a long short-term memory network model;

[0030] NB3: The energy consumption data and environmental data acquired and processed in real time are used as the input of each base model to obtain the output of each base model, and the output of each base model is weighted averaged to obtain the corresponding energy optimization result.

[0031] Furthermore, it also includes a positioning module electrically connected to the energy management module and the multi-dimensional display module, which is used to obtain the location information of the park equipment / facility and send the location information to the multi-dimensional display module.

[0032] Furthermore, the multi-dimensional display module displays the energy optimization results by equipment, region and time period, and displays the data of the energy optimization results by charts and dashboards.

[0033] A smart park big data comprehensive management and control method uses any of the smart park big data comprehensive management and control systems described above.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] First, the present invention acquires the energy consumption data and external environment data of the park in real time, so that it can predict the future energy consumption based on the actual data and output the energy optimization results, so that the park can dynamically adjust the energy use strategy to reduce energy waste. At the same time, it can also formulate personalized energy-saving plans according to the behavior habits and needs of different users, which helps to improve users' energy-saving awareness and encourage users to actively participate in energy-saving actions through energy-saving incentive activities;

[0036] Second, the linkage unit of the present invention can coordinate the intelligent linkage mode between different devices in the park according to the energy scheduling plan and the personalized energy-saving plan of the park, thereby not only optimizing the energy utilization efficiency, but also improving the service life and reliability of the equipment;

[0037] Thirdly, the present invention not only helps balance the load of the power grid but also reduces the energy cost of the park by identifying the time period of low power supply and storing electric energy during this period;

[0038] Fourthly, the present invention improves the accuracy and robustness of prediction by constructing multiple base models and using meta-models for integrated learning, so as to better cope with complex and changeable energy consumption and environmental conditions and improve the effect of energy optimization;

[0039] Fifth: The multi-dimensional display module of the present invention can not only display the energy optimization results from multiple dimensions such as equipment, area and time period, but also can intuitively present data through charts and dashboards. At the same time, the positioning module can obtain the location information of the park equipment and send it to the multi-dimensional display module, so as to facilitate management personnel to quickly locate the equipment and perform maintenance and repair in time. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a system block diagram of the smart park big data comprehensive management and control system of the present invention;

[0041] Figure 2 Schematic diagram of the working process of the energy storage module of the present invention;

[0042] Figure 3 A schematic diagram of a process for obtaining energy optimization results according to the present invention;

[0043] Figure 4 A flow chart of another method for obtaining energy optimization results according to the present invention. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] With the rapid economic development and social progress, industrial parks have become an important force in promoting local economic development. At the same time, the continuous expansion of the park scale and the increase in the number of enterprises entering the park have also led to an increasing energy consumption, making it difficult for traditional energy management models to meet the needs of modern industrial parks for high efficiency, intelligence and sustainable development.

[0046] Example 1

[0047] refer to Figure 1, this embodiment provides a smart park big data comprehensive management and control system, which includes a data module, an energy management module, a multi-dimensional display module and a park service module. The data module is used to obtain external data for other modules to call data, that is, to obtain the energy consumption data of each facility in the park, the energy consumption data of each user and the external environment data of the park. Furthermore, the energy management module is used to receive energy consumption data and external environment data, and based on the energy consumption data and external environment data, combined with the energy prediction model, predict the corresponding energy optimization results, and send the energy optimization results to the multi-dimensional display module and the park service module. Specifically, the multi-dimensional display module can display the energy optimization results in multiple dimensions. The park service module can make specific adjustments to the operation of the park equipment / facilities, users, heating or cooling systems according to the energy optimization results.

[0048] In this embodiment, the data module can obtain the energy consumption data of the park and the external environment data of the park in real time, and process the energy consumption data and the external environment data of the park. Specifically, the energy consumption data and the external environment data of the park obtained by the data module in this embodiment need to be further cleaned and standardized to facilitate subsequent unified storage, and also facilitate the unified processing of different data by other modules. Further, in the process of data cleaning, it includes removing invalid data and filling missing values, wherein removing invalid data means removing erroneous data caused by sensor failure or other reasons, and filling missing values ​​means filling missing data points by interpolation or other statistical methods. In the process of data standardization, it includes unit conversion and format unification, wherein unit conversion means that all energy consumption data use the same unit, and format unification means converting data from different devices into a consistent format. It is worth noting that the data cleaning and standardization in this embodiment are both conventional technical means, so they will not be specifically explained in this embodiment.

[0049] Furthermore, the data module can obtain the energy consumption data of each device in the park in real time through different data acquisition devices. For example, it can measure the consumption of energy such as electricity, water and natural gas in the park by installing smart meters, smart water meters and smart gas meters in various places in the park. At the same time, environmental sensors such as temperature, humidity and light, as well as flow meters, pressure gauges and other equipment can be installed in the park to monitor the status of the environment and equipment. At the same time, the energy consumption of connected devices can be monitored in real time by monitoring IoT devices, such as real-time monitoring through smart sockets and smart switches. In other words, after the data module monitors the equipment in the park through smart meters / water meters / gas meters, sensors and IoT devices, it can record the energy consumption behavior of users through applications on user terminal devices (such as: mobile devices such as smartphones and tablets). At the same time, fixed equipment in the park (such as air conditioners, elevators, lighting systems, etc.) can upload their energy consumption data through embedded systems or by connecting to the Internet. It is worth noting that the energy consumption data can also be classified according to the different users in the park. That is, the energy consumption data corresponding to each user can be obtained through the energy consumption monitoring of the equipment, so as to predict the corresponding personalized energy-saving plan according to the energy consumption usage of different users.

[0050] Furthermore, the data module can also collect environmental parameters such as temperature, humidity, wind speed, wind direction and rainfall in the park in real time through weather stations set up inside or around the park. It is also possible to obtain a wide range of meteorological information, such as cloud coverage, surface temperature, etc., through satellite remote sensing technology. It is also possible to connect to the API interface of a third-party meteorological service provider and call the meteorological data of the third-party meteorological service provider to obtain accurate weather forecast data. At the same time, the data module will also send the obtained external environmental data of the park to the energy management module, so that the energy prediction model in the energy management module can intelligently adjust the startup temperature threshold of the heating or cooling system according to the external environmental data of the park, so that energy saving can be achieved while meeting the comfort level.

[0051] In this embodiment, the energy management module can determine the energy optimization result based on the above-processed energy consumption data and the park's external environment data, combined with the corresponding energy prediction model, and send the energy optimization result to the multi-dimensional display module, and perform various displays through data visualization, interactive interface, multi-dimensional display, mobile / Web application, etc. At the same time, the energy optimization result is also sent to the park service module to adjust the working status and operating parameters of the park equipment / facilities, users, heating or cooling systems.

[0052] Specifically, in this embodiment, the multi-dimensional display module can display data visualization through charts and dashboards. Chart display can use bar charts, line charts, pie charts and other charts to display energy consumption forecast results, actual energy consumption data and environmental data. Dashboard display is to display pipe indicators such as current energy consumption level, energy storage status and equipment operation status through dashboard charts or progress bars.

[0053] Furthermore, the multi-dimensional display module in this embodiment can be divided into real-time monitoring and historical trend analysis when performing interactive interface display. Real-time monitoring provides a real-time monitoring interface to display the current energy consumption status and equipment operation status, etc. Historical trend analysis provides trend analysis of historical data to help users understand the changing trend of energy consumption.

[0054] Furthermore, the multi-dimensional display module in this embodiment can be divided into device display, area display and time period display when performing multi-dimensional display. The device display is to display the energy consumption trend of each device according to the energy consumption of different devices. The area display is to display the energy consumption distribution of each area according to the energy consumption of different areas in the park. The time period display is to display the energy consumption changes by time period, such as by hours, days, cycles, etc.

[0055] Furthermore, the multi-dimensional display module in this embodiment can be divided into mobile application and Web application display when performing mobile / Web application display. The mobile application can allow users to view energy consumption information anytime and anywhere through mobile application software. The Web application can build a Web application platform to facilitate park managers and users to access energy consumption data through a browser.

[0056] Specifically, the energy optimization results in this embodiment include energy scheduling plans, park personalized energy-saving plans, user personalized energy-saving plans, expected benefits after energy-saving plans, and working parameters of heating or cooling systems. At the same time, the park service module in this embodiment operates specifically according to the energy optimization results. The park service module in this embodiment includes a linkage unit, a participation unit, and an adjustment unit.

[0057] Furthermore, the linkage unit is used to determine the intelligent linkage mode between different devices in the park according to the energy scheduling plan and the personalized energy-saving plan of the park. For example, when the outdoor temperature is higher than the indoor set temperature (for example, 25°C), if the window is closed, the air conditioner is turned on. When the outdoor temperature is lower than the indoor set temperature (for example, 25°C), when the window is open, the air conditioner is turned off. That is, through the intelligent control center, the switch status of the window is monitored, and the working mode of the air conditioner is adjusted according to the outdoor temperature. Similarly, when the human body sensor detects human activity, the lighting equipment in the corresponding area is turned on. When the human body sensor does not detect activity within the preset time (for example, within 120s), the lighting equipment is automatically turned off. That is, through the intelligent lighting system, it is linked with the human body sensor to automatically control the switch of the light.

[0058] Further, the participating unit is used to formulate energy-saving incentive activities according to the user's personalized energy-saving plan, which includes energy-saving competitions, reward points, energy-saving challenges, energy-saving funds and energy-saving certification. Among them, energy-saving competitions are to hold energy-saving competitions regularly, rank users according to their energy-saving contributions, and give material rewards or honorary certificates to the top three users. At the same time, the competition can also be divided into individual competitions and team competitions to encourage group cooperation and jointly achieve energy-saving goals. Reward points means that users can get points by implementing energy-saving plans, and points can be exchanged for physical rewards such as small gifts and coupons. Energy-saving challenges are to issue a series of energy-saving challenge tasks, such as reducing energy consumption by a certain proportion (such as 20%) within a month, and users who complete the tasks can get rewards. Energy-saving funds are to set up energy-saving funds, invest the saved energy costs into the funds, and use them to support the implementation of energy-saving and emission reduction projects or distribute them to participants as rewards. Energy-saving certification is to issue energy-saving certification to enterprises or individuals that meet certain energy-saving standards, such as "Energy-Saving Star", "Green Ambassador" and other titles, and publicize them in the park to enhance their social influence.

[0059] Furthermore, the adjustment unit is used to adjust the start-up temperature threshold of the heating or cooling system according to the working parameters of the heating or cooling system. That is, the real-time working status of the heating or cooling system is monitored, including parameters such as temperature and pressure, and the user's comfort feedback information is recorded. Based on the current environmental conditions and user feedback, the start-up temperature threshold is determined. In other words, by dynamically adjusting the start-up temperature of the heating or cooling system, it is ensured that both the user's comfort requirements are met and excessive energy consumption is avoided.

[0060] refer to Figure 3 , the method for obtaining the energy optimization result in this embodiment is as follows:

[0061] Step NA1: Construct feature engineering through energy consumption data and environmental data processed by the data module, and the feature engineering includes time window features, seasonal features and weather features. The time window features are divided into historical energy consumption data (i.e., the average energy consumption per hour in the past week) and time features (i.e., extracting time features such as hours, weeks, and months). Seasonal features are divided into seasons (i.e., judging the current season based on the current date) and holidays (i.e., whether it is a holiday). Weather features are divided into real-time weather data (i.e., temperature, humidity, wind speed, etc.) and weather types (i.e., sunny, rainy, snowy, etc.).

[0062] That is to say, the feature engineering bits in this embodiment are: time window features (i.e., using the average energy consumption of the past 24 hours and the daily peak energy consumption of the past week), seasonal features (i.e., recording the date corresponding to today and whether it is a holiday), and weather features (i.e., the current outdoor temperature is 20°C, the humidity is 60%, and the wind speed is 5km / h).

[0063] Step NA2: Based on the feature engineering extracted in step NA1, multiple base models are constructed and a meta-feature matrix is ​​obtained, wherein the multiple base models include a random forest model, a support vector machine model, a gradient boosting tree model and a long short-term memory network model. The details are as follows:

[0064] Step NA2.1: Based on the feature engineering extracted in step NA1, multiple base models are constructed and each base model is trained by cross-validation method. The random forest model is used to predict the energy consumption in the next 24 hours, the support vector machine model is used to predict the peak energy consumption tomorrow, the gradient boosting tree model is used to predict the average daily energy consumption in the next week, and the long short-term memory network model is used to predict the energy consumption in the next week.

[0065] Furthermore, the feature engineering extracted in the above step NA1 is divided into a training set and a validation set, that is, each base model is trained using the training set, and each base model after training is validated using the validation set, and data prediction can be performed after the validation passes. It is worth noting that the training and validation of each base model here is a conventional technical means, so it is not specifically described in this embodiment.

[0066] Step NA2.2: Take the prediction results corresponding to each base model training as meta-features, and combine the meta-features corresponding to each base model to form a meta-feature matrix, where each column in the meta-feature matrix is ​​the prediction value of a base model, and each row in the meta-feature matrix is ​​a data point.

[0067] Step NA3: construct at least one meta-model, and use the meta-feature matrix as the input of the meta-model to train the meta-model, wherein the meta-model includes a linear regression model, a logistic regression model, and a neural network model.

[0068] Step NA4: The energy consumption data and environmental data acquired and processed in real time are used as the input of the base model to obtain the real-time meta-feature matrix. Meanwhile, the real-time meta-feature matrix is ​​used as the input of the trained meta-model to output the corresponding energy optimization result.

[0069] In the specific implementation process, today is Wednesday, the outdoor temperature is 20°C, the humidity is 60%, and the wind speed is 5km / h. The average energy consumption in the past 24 hours is 1000 kWh, and the daily average energy consumption peak in the past week is 1200 kWh. Based on this, the energy optimization results corresponding to the park can be obtained, which are as follows:

[0070] Through the above data, the feature engineering is time window feature (the average energy consumption in the past 24 hours is 1000 kWh), seasonal feature (today is Wednesday, not a holiday), and weather feature (the current outdoor temperature is 20℃, the humidity is 60%, and the wind speed is 5km / h).

[0071] At this time, through this feature engineering, the prediction results corresponding to the above base models are:

[0072] The random forest model predicts that the energy consumption for the next 24 hours will be 1050 kWh.

[0073] The support vector machine model predicts that the peak energy consumption tomorrow will be 1250 kWh.

[0074] The gradient boosting tree model predicts that the average daily energy consumption for the next week will be 1030 kWh.

[0075] The LSTM network model predicts that the average daily energy consumption for the next week will be 1040 kWh.

[0076] That is to say, the meta-feature matrix is ​​[1050, 1250, 1030, 1040]. The meta-model is constructed based on the meta-feature matrix, and the obtained results are as follows:

[0077] The linear regression model predicts that the energy consumption for the next 24 hours will be 1035 kWh.

[0078] The neural network model predicts that the energy consumption in the next 24 hours will be 1045 kWh.

[0079] Example 2

[0080] refer to Figure 2This embodiment provides a smart park big data integrated management and control system, the basic composition of which is the same as that of embodiment 1, except that the smart park big data integrated management and control system further includes an energy storage module, which is electrically connected to the linkage unit and the external power supply. Specifically, the energy storage module can store electric energy according to the level of electricity price, that is, store electric energy when the electricity price is low, and release electric energy when the electricity price is high, so as to achieve the purpose of peak-shifting use and reduce the energy input cost of the park.

[0081] In this embodiment, the specific implementation process of the energy storage module is as follows:

[0082] Step M1: Extract characteristic values ​​from historical power consumption data and electricity price data, the characteristic values ​​including average electricity price, electricity price fluctuation range, electricity price standard deviation and electricity price change rate.

[0083] In the specific embodiment, the electricity price data is as follows:

[0084] Table 1: Electricity price data for the past month

[0085] time Electricity price (yuan / kWh) t1 0.5 T2 0.45 … … tn 0.6

[0086] That is, according to Table 1 above, we can know that the average electricity price is 0.5 yuan / kWh and the standard deviation of the electricity price is 0.05 yuan / kWh.

[0087] Step M2: Determine the power supply valley time period according to the characteristic value obtained in step M1, specifically:

[0088] T low ={t|P(t)≤α·(μ+β·σ)}

[0089] Where: T low is the power supply off-peak period, t is the power supply time, P(t) is the electricity price corresponding to the power supply time t, α is a constant, μ is the average electricity price, β is a constant, and σ is the standard deviation of the electricity price.

[0090] In this embodiment, the constant α is set to 0.9, and the constant β is set to 1. According to the above formula, the electricity price threshold for the low power supply period is 0.495 yuan / kWh. That is, when the electricity price is not higher than 0.495 yuan / kWh, the period is the low power supply period.

[0091] Step M3: Determine whether the current time period is in the power supply valley time period. If it is in the power supply valley time period, electrically connect the energy storage battery in the energy storage module to the external power supply. If it is not in the power supply valley time period, electrically connect the energy storage battery in the energy storage module to the linkage unit.

[0092] That is to say, when the electricity price in the current time period is not higher than 0.495 yuan / kWh, it is a low power supply period, that is, the energy storage battery in the energy storage module is electrically connected to the external power supply, that is, the energy storage battery is charged at this time.

[0093] When the electricity price in the current time period is higher than 0.495 yuan / kWh, it is not a low power supply period, that is, the energy storage battery in the energy storage module is electrically connected to the linkage unit, that is, the energy storage battery is discharged at this time.

[0094] Example 3

[0095] refer to Figure 4 This embodiment provides a smart park big data integrated management and control system, the basic composition of which is the same as that of Embodiment 1, except that the energy optimization result in this embodiment is obtained in the following manner:

[0096] Step NB1: construct feature engineering through the energy consumption data and environmental data processed by the data module, and the feature engineering includes time window features, seasonal features and weather features. The time window feature is the energy consumption data in the past period of time (i.e. the energy consumption data of the past day), the seasonal feature is the season of the year (spring, summer, autumn and winter), which day of the week (weekday or weekend), and the weather features are divided into outdoor temperature, humidity and wind speed.

[0097] That is to say, the feature engineering bits in this embodiment are: time window features (i.e., the average energy consumption in the past 24 hours is 1000 kWh), seasonal features (i.e., today is Wednesday, not a holiday), and weather features (i.e., the current outdoor temperature is 20°C, the humidity is 60%, and the wind speed is 5 km / h).

[0098] Step NB2: Build multiple base models, and divide feature engineering into training sets and validation sets, that is, 70% of feature engineering is training sets and 30% is validation sets. Each base model is trained with the training set, and each trained base model is validated with the validation set. Multiple base models include random forest model, support vector machine model, gradient boosting tree model and long short-term memory network model.

[0099] Step NB3: The energy consumption data and environmental data acquired and processed in real time are used as the input of each base model to obtain the output of each base model, and the output of each base model is weighted averaged to obtain the corresponding energy optimization result.

[0100] In the process of the specific embodiment, the current outdoor temperature is 21° C., the humidity is 62%, the wind speed is 6 km / h, and the average energy consumption in the past 24 hours is 1050 kWh.

[0101] At this time, through this feature engineering, the prediction results corresponding to the above base models are:

[0102] The random forest model predicts that the energy consumption for the next 24 hours will be 1050 kWh.

[0103] The support vector machine model predicts that the peak energy consumption tomorrow will be 1200 kWh.

[0104] The gradient boosting tree model predicts that the average daily energy consumption for the next week will be 1030 kWh.

[0105] The LSTM network model predicts that the average daily energy consumption for the next week will be 1040 kWh.

[0106] At the same time, the weight of each model is set to 0.25, and the final energy optimization result is 1080 kWh.

[0107] Example 4

[0108] This embodiment provides a smart park big data integrated management and control system, the basic composition of which is the same as that of embodiment 1, except that the smart park big data integrated management and control system further includes a positioning module, which is electrically connected to the energy management module and the multi-dimensional display module. Specifically, the positioning module in this embodiment can obtain the location information of each device / facility in the park through various methods such as RFID tags and card readers, Bluetooth beacons, Wi-Fi positioning, GPS / GNSS, and ultra-wideband technology.

[0109] Furthermore, when the multi-dimensional display module displays energy optimization results, energy consumption data, and external environment data of the park, users can specifically view individual data with errors or large changes, and match the data with the corresponding equipment / facilities. At the same time, the positioning module can be used to directly locate the specific location in the park, so that staff can be quickly dispatched for inspection and maintenance. In other words, the setting of the positioning module can not only enable park managers to quickly discover and solve equipment problems, but also reduce energy waste caused by equipment failures, and improve the efficiency and management level of park operations.

[0110] Example 5

[0111] This embodiment provides a method for comprehensive management and control of big data in a smart park. This method for comprehensive management and control of big data in a smart park is used to implement the above embodiments and preferred implementation methods, and those that have been explained will not be repeated.

[0112] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is limited by the attached embodiments and their equivalents.

Claims

1. A smart park big data comprehensive management and control system, characterized in that: Included are: A data module, which acquires the energy consumption data of the park and the external environment data of the park in real time, and processes the energy consumption data and the environmental data, wherein the energy consumption data includes the personal energy consumption data of the user; The energy management module uses the processed energy consumption data and environmental data as the input of the prediction model, and outputs the energy optimization results, which include energy scheduling solutions, personalized energy-saving solutions for parks, personalized energy-saving solutions for users, and working parameters of heating or cooling systems. At the same time, the energy optimization results are sent to the park service module and the multi-dimensional display module. The multi-dimensional display module is used to display the energy optimization results in multiple dimensions. The park service module includes: A linkage unit, which determines the intelligent linkage mode between different devices in the park according to the energy scheduling plan and the park's personalized energy-saving plan; Participating units formulate energy-saving incentive activities according to the user's personalized energy-saving plan, and the energy-saving incentive activities include energy-saving competitions, reward points, energy-saving challenges, energy-saving funds and energy-saving certification; The adjusting unit adjusts the starting temperature threshold of the heating or cooling system according to the working parameters of the heating or cooling system.

2. According to claim 1, a smart park big data comprehensive management and control system is characterized in that: It also includes an energy storage module electrically connected to the linkage unit and the external power supply, and stores electric energy by identifying the power supply time period, specifically including: M1: Extract characteristic values ​​from historical power consumption data and electricity price data, the characteristic values ​​include average electricity price, electricity price fluctuation range, electricity price standard deviation and electricity price change rate; M2: Determine the power supply valley time period according to the characteristic value, specifically: T low ={t|P(t)≤α·(μ+β·σ)} Where: T low is the power supply off-peak period, t is the power supply time, P(t) is the electricity price corresponding to the power supply time t, α is a constant, μ is the average electricity price, β is a constant, and σ is the standard deviation of the electricity price; M3: Determine whether the current time period is in the power supply valley time period. When it is in the power supply valley time period, electrically connect the energy storage battery in the energy storage module to the external power supply. When it is not in the power supply valley time period, electrically connect the energy storage battery in the energy storage module to the linkage unit.

3. According to claim 1, a smart park big data comprehensive management and control system is characterized in that: The energy consumption data and environmental data are processed by data cleaning and standardization.

4. According to claim 1, a smart park big data comprehensive management and control system is characterized in that: The energy optimization result is obtained by: NA1: construct feature engineering through the processed energy consumption data and environmental data, and the feature engineering includes time window features, seasonal features and weather features; NA2: According to the feature engineering, multiple base models are constructed and a meta-feature matrix is ​​obtained. The multiple base models include a random forest model, a support vector machine model, a gradient boosting tree model and a long short-term memory network model; NA3: construct at least one meta-model, and use the meta-feature matrix as an input of the meta-model to train the meta-model, wherein the meta-model includes a linear regression model, a logistic regression model, and a neural network model; NA4: The energy consumption data and environmental data acquired and processed in real time are used as the input of the base model to obtain a real-time meta-feature matrix. Meanwhile, the real-time meta-feature matrix is ​​used as the input of the trained meta-model to output the corresponding energy optimization result.

5. A smart park big data integrated management and control system according to claim 4, characterized in that: The meta-feature matrix is ​​obtained, specifically including: NA2.1: Based on the feature engineering, multiple base models are constructed, and each base model is trained by a cross-validation method; NA2.2: The prediction results corresponding to the base model training are used as meta-features, and the meta-features corresponding to each base model are combined to form a meta-feature matrix.

6. According to claim 1, a smart park big data integrated management and control system is characterized in that: The method for obtaining the energy optimization result also includes: NB1: construct feature engineering through the processed energy consumption data and environmental data, and the feature engineering includes time window features, seasonal features and weather features; NB2: construct multiple base models, divide the feature engineering into a training set and a validation set, train each base model through the training set, and validate each trained base model through the validation set, wherein the multiple base models include a random forest model, a support vector machine model, a gradient boosting tree model, and a long short-term memory network model; NB3: The energy consumption data and environmental data acquired and processed in real time are used as the input of each base model to obtain the output of each base model, and the output of each base model is weighted averaged to obtain the corresponding energy optimization result.

7. The smart park big data integrated management and control system according to claim 1 is characterized in that: It also includes a positioning module electrically connected to the energy management module and the multi-dimensional display module, which is used to obtain the location information of the equipment / facility in the park and send the location information to the multi-dimensional display module.

8. A smart park big data integrated management and control system according to claim 1 or 7, characterized in that: The multi-dimensional display module displays the energy optimization results by equipment, region and time period, and displays the data of the energy optimization results by charts and dashboards.

9. A comprehensive management and control method for big data in a smart park, characterized in that: The smart park big data comprehensive management and control system described in any one of claims 1-8 is used.

Images