Cooling and heating control method and system based on energy conservation and energy storage

Abstract

The invention discloses a cooling and heating control method and system based on energy conservation and energy storage, and relates to the technical field of cooling and heating control. A cooling and heating control system based on energy conservation and energy storage comprises an energy supply data acquisition module and an energy supply strategy optimization module. According to the method, on the basis of historical load data, meteorological information and equipment operation characteristics, future cooling and heating load requirements are accurately predicted, and multi-dimensional balance among operation cost, switching cost and user comfort is realized in combination with a time-of-use electricity price mechanism; rigid loads and adjustable loads are flexibly identified through a judgment mechanism of a clear energy supply demand and a fuzzy energy supply demand, and the accuracy and flexibility of scheduling response are improved; an energy supply strategy is dynamically adjusted through a rolling optimization algorithm, the cold and heat storage capacity is utilized to the maximum degree on the premise that the user requirements are met, power consumption in the peak period is reduced, the service life of equipment is prolonged, and the overall energy efficiency level is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cooling and heating control, and particularly relates to a cooling and heating control method and system based on energy-saving and energy storage. BACKGROUND

[0002] The existing regional energy supply system usually relies on large-capacity main machines to meet the peak of cooling and heating load, which has large initial investment and low equipment utilization rate. The cooling and heating control is usually operated according to real-time load, ignoring the time-of-use electricity price and equipment efficiency change, and the overall energy efficiency is limited. At the same time, the cooling and heating load has large seasonal difference, and if there is lack of intelligent scheduling, it is easy to cause imbalance between supply and demand and inefficient operation, and extreme climate change may also cause system performance fluctuation and energy waste.

[0003] Therefore, there is a need for a cooling and heating control method based on energy-saving and energy storage, which takes load prediction, cost optimization and rolling scheduling as the core, realizes efficient, safe and economic intelligent energy supply response, and solves the above technical problems. SUMMARY

[0004] The present application aims to provide a cooling and heating control method and system based on energy-saving and energy storage, which takes energy-saving and energy storage as the premise and adapts to different external climate changes.

[0005] A cooling and heating control method based on energy-saving and energy storage, comprising the following steps: Setting an energy supply future time domain; obtaining energy supply parameters of energy storage energy supply equipment and current cooling and heating load parameters before the start of the energy supply future time domain; the energy supply parameters include hourly load prediction values, time-of-use electricity price table and operation efficiency performance model generated based on historical data and weather information; based on the energy supply parameters and the current cooling and heating load parameters, the prediction is carried out to obtain energy supply demand determination and predicted cooling and heating load parameter sequence; the energy supply demand determination includes clear energy supply demand and fuzzy energy supply demand; Obtaining energy storage energy supply equipment performance model and energy storage tank state data; based on the predicted cooling and heating load parameter sequence and the pre-trained energy-saving cost model, the cost analysis is carried out to obtain direct operation cost, switching decision cost and comfort deviation cost; According to the direct operation cost, the switching decision cost and the comfort deviation cost, the rolling optimization is solved, when the energy supply demand determination is the clear energy supply demand, the energy storage energy supply strategy obtained by the rolling optimization is used to control the energy storage energy supply equipment; when the energy supply demand determination is the fuzzy energy supply demand, the optimal decision selection is carried out based on the obtained energy storage energy supply strategy and the comfort deviation cost to obtain the optimal energy storage energy supply strategy, and the optimal energy storage energy supply strategy is used to control the energy storage energy supply equipment.

[0006] As a preferred technical solution of the present application, the specific steps of obtaining the energy supply parameters of the energy storage energy supply equipment and the current cooling and heating load parameters include: acquire actual cold and heat load data corresponding to time stamp and calendar sequence containing special events in the past n years, acquire meteorological parameters corresponding to time stamp in the future time domain, train the actual cold and heat load data and calendar sequence as training features to obtain a trained cold and heat load prediction model; input the meteorological parameters into the trained cold and heat load prediction model for prediction to obtain hourly load prediction values; acquire a running state data set of the energy storage and supply equipment, wherein each piece of data in the running state data set contains equipment load rate, equipment input power consumption and key running boundary temperature; construct a running efficiency performance model based on the running state data set and a regression algorithm.

[0007] As a preferred technical solution of the present application, the specific steps of prediction based on the supply energy parameters and the current cold and heat load parameters include: perform feature fusion based on the current cold and heat load parameters and the hourly load prediction values to obtain a predicted cold and heat load parameter sequence in the future time domain of supply energy, which is sorted by time steps; perform clustering on the predicted cold and heat load parameter sequence to obtain a predicted low load cluster, a predicted medium load cluster and a predicted high load cluster, and set the minimum value of the predicted cold and heat load parameters in the predicted high load cluster as a supply energy determination threshold; if the number of time steps in which the predicted cold and heat load parameters exceed the supply energy determination threshold accounts for more than M% of the total time steps in the predicted cold and heat load parameter sequence, and there are at least P time periods in which all the predicted cold and heat load parameters in each time period exceed the supply energy determination threshold for more than T time steps, then the supply energy demand in the future time domain is determined as clear supply energy demand; otherwise, the supply energy demand is determined as fuzzy supply energy demand.

[0008] As a preferred technical solution of the present application, the specific steps of cost analysis based on the predicted cold and heat load parameter sequence and the pre-trained energy saving cost model include: perform calculation based on the predicted cold and heat load parameter sequence and the running efficiency performance model to obtain equipment predicted power consumption; perform calculation based on the equipment predicted power consumption and a time-of-use electricity price table to obtain direct running cost; When the clear energy supply demand and the current cold and heat load parameter characteristics are opposite, the switching decision cost is calculated; otherwise, the switching decision cost is set as a standard value; the specific steps for calculating the switching decision cost are that the state data of the energy storage water tank contains a reference water temperature, an average water temperature and a total water capacity; the energy of the state data of the energy storage water tank is calculated by using a physical method to obtain a switching energy internal consumption cost; and the total time required to reach the clear energy supply demand is determined based on a pre-trained energy saving cost model to obtain a switching capacity time cost; and the switching decision cost is obtained based on the switching energy internal consumption cost and the switching capacity time cost; The optimal efficiency operating point and the best load rate of the energy storage energy supply device are determined according to the operating efficiency performance model; the device load rate corresponding to the predicted cold and heat load parameter sequence is calculated; the efficiency penalty factor is generated based on the best load rate and the device load rate; and the comfort deviation cost is obtained by calculating based on the efficiency penalty factor and the direct operating cost.

[0009] As a preferred technical solution of the present application, when the energy supply demand is determined as the clear energy supply demand, a target optimization function is constructed by combining the direct operating cost, the switching decision cost and the comfort deviation cost; the energy storage energy supply strategy is obtained by solving the rolling optimization algorithm with the target optimization function as the optimization target and the energy storage energy supply device performance model and the energy storage water tank state data as the constraint conditions.

[0010] As a preferred technical solution of the present application, when the energy supply demand is determined as the fuzzy energy supply demand, the current energy storage energy supply strategy is obtained; the predicted comprehensive cost target function values generated by executing the current energy storage energy supply strategy and the energy storage energy supply strategy are calculated respectively; and the optimal energy storage energy supply strategy is defined based on the predicted comprehensive cost target function values.

[0011] An energy-saving energy storage-based cooling and heating control system, comprising: An energy supply data acquisition module, comprising a load prediction unit; the load prediction unit is used for setting an energy supply future time domain; energy supply parameters and current cold and heat load parameters of an energy storage energy supply device are acquired before the start of the energy supply future time domain; the energy supply parameters comprise hourly load prediction values, time-of-use electricity price tables and operating efficiency performance models generated based on historical data and meteorological information; energy supply demand determination and predicted cold and heat load parameter sequences are obtained by prediction based on the energy supply parameters and the current cold and heat load parameters; the energy supply demand determination comprises clear energy supply demand and fuzzy energy supply demand; An energy supply strategy optimization module, comprising a cost analysis unit and an energy supply optimization unit; the cost analysis unit is used for acquiring an energy storage energy supply device performance model and energy storage water tank state data; direct operating cost, switching decision cost and comfort deviation cost are obtained by cost analysis based on the predicted cold and heat load parameter sequences and a pre-trained energy saving cost model; The energy supply optimization unit is used for solving the rolling optimization according to the direct operation cost, the switching decision cost and the comfort deviation cost, and when the energy supply demand is determined as clear energy supply demand, the energy storage energy supply equipment is controlled by using the energy storage energy supply strategy obtained by solving the rolling optimization; when the energy supply demand is determined as fuzzy energy supply demand, the optimal decision selection is made based on the obtained energy storage energy supply strategy and the comfort deviation cost, the optimal energy storage energy supply strategy is obtained, and the energy storage energy supply equipment is controlled by using the optimal energy storage energy supply strategy.

[0012] The present application has the following advantages: 1、The present application accurately predicts future cold and heat load demand on the basis of historical load data, weather information and equipment operation characteristics, and realizes multi-dimensional balance between operation cost, switching cost and user comfort by combining with time-of-use electricity price mechanism; through the determination mechanism of clear energy supply demand and fuzzy energy supply demand, rigid load and adjustable load are flexibly identified, and the accuracy and flexibility of dispatching response are improved; through the rolling optimization algorithm, the energy supply strategy is dynamically adjusted, the cold and heat storage capacity is maximized under the premise of meeting user demand, the peak period power consumption is reduced, the equipment service life is prolonged, and the overall energy efficiency level is improved.

[0013] 2、The present application distinguishes clear energy supply and fuzzy energy supply scenes by introducing cold and heat load clustering analysis and energy supply demand determination mechanism, realizes intelligent identification of load intensity and continuity, so that the dispatching strategy is more targeted and forward-looking; a multi-objective cost model composed of direct operation cost, switching decision cost and comfort deviation cost is constructed, so that the control strategy not only pursues the minimum energy consumption, but also considers the equipment switching efficiency and user experience, and realizes dynamic optimization of comprehensive cost. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 A structure schematic view of a cooling and heating control system based on energy saving and energy storage adopted by the embodiment of the present application. DETAILED DESCRIPTION

[0015] In order to make the person skilled in the art better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application.

[0016] Embodiment 1, a cooling and heating control method based on energy saving and energy storage, comprising the following steps: Setting an energy supply future time domain; obtaining energy supply parameters and current cold and heat load parameters of the energy storage energy supply device before the start of the energy supply future time domain; the energy supply parameters include hourly load prediction values, time-of-use electricity price tables and operation efficiency performance models generated based on historical data and meteorological information; based on the energy supply parameters and the current cold and heat load parameters, a prediction is made to obtain an energy supply demand determination and a predicted cold and heat load parameter sequence; the energy supply demand determination includes clear energy supply demand and fuzzy energy supply demand; The energy supply future time domain is a time range in which the system determines to perform energy supply scheduling and management in the future for a certain period of time, which is usually in units of hours, for example, a time sequence of 24 hours or 48 hours in the future, which is set by professional technicians artificially; The specific steps of obtaining the energy supply parameters and the current cold and heat load parameters of the energy storage energy supply device include: Obtaining hourly actual cold and heat load data corresponding to the time stamp for at least n complete operation years and a calendar sequence corresponding to the time stamp and containing special events; obtaining hourly meteorological parameters corresponding to the time stamp in the energy supply future time domain; training the actual cold and heat load data and the calendar sequence as training features to obtain a trained cold and heat load prediction model; The input of the above steps is at least n complete operation years of historical hourly actual cold and heat load data corresponding to the time stamp, which is extracted from a historical database, and a calendar sequence containing special events (such as statutory holidays, major event days, and weather anomaly days) is also input; another input is hourly meteorological parameters from a meteorological service agency, covering the future energy supply time domain and also corresponding to the time stamp, mainly including outdoor dry-bulb temperature, wet-bulb temperature, solar radiation, wind speed and other key environmental variables; The output is a structured feature data set for model training, which is represented as a multi-table, each row representing a historical hour, and the columns containing the time stamp, actual cold and heat load value, and multiple features derived from the calendar sequence and the time stamp, such as the day of the week, whether it is a holiday, whether it is a special event day, etc.; data extraction, conversion and loading processes are used to collect data through database queries and application program interface calls, and data cleaning, alignment and feature engineering are performed, providing rich and high-quality training samples for training a high-precision prediction model that can deeply understand the operation rules of the building itself and external influencing factors, which is the fundamental prerequisite for precise prediction; a supervised machine learning algorithm is used to build a cold and heat load prediction model, and the feature data set is trained through repeated iterations to minimize the error between the model prediction value and the historical true value until a trained cold and heat load prediction model is obtained.

[0017] The hourly meteorological parameters are input into the trained cold and heat load prediction model for prediction to obtain hourly load prediction values; the hourly load prediction values are a one-dimensional time sequence array, each element of which contains a future time stamp and a corresponding predicted load value.

[0018] An operating state data set of the energy storage and supply device is obtained, wherein each piece of data of the operating state data set contains a device load rate, a device input power consumption, and a key operating boundary temperature; the device load rate refers to the ratio of the current actual operating load of the device to the rated maximum design load thereof, is usually expressed in percentage, is used for describing the operating intensity of the device, and is an important basis for judging whether the device is in partial load or full load operation; for example, if a water source heat pump with a rated refrigeration capacity of 1000 kW currently outputs a refrigeration capacity of 600 kW, the load rate thereof is 60%; the device input power consumption refers to the electric power consumed by the device at the time for maintaining the current output, is usually in kilowatts (kW), includes the total power consumption of a compressor, a motor, a water pump, and the like, and is a direct index for measuring the operating cost: the power consumption multiplied by the electricity price is the electricity charge; the key operating boundary temperature reflects the inlet and outlet water temperatures of a key heat exchange part in the operation of the device, usually includes the following types: when the energy storage and supply device performs cooling, the chilled water outlet temperature and the cooling water inlet temperature; when the energy storage and supply device performs heating, the hot water outlet temperature and the source side fluid inlet temperature, which are directly related to the thermodynamic cycle process of the heat pump or the water chiller.

[0019] An operating efficiency performance model is constructed based on the operating state data set and a regression algorithm; the construction steps are as follows: the input of the operating efficiency performance model is the key operating boundary temperature and the device load rate, the output is the device input power consumption, the input parameters can be mapped to the operating efficiency of the device, and the operating efficiency performance model can be constructed by using a polynomial regression, a decision tree regression, or a neural network, fits the efficiency performance of the device under different operating conditions, and obtains the operating efficiency performance model.

[0020] The specific steps of prediction based on the energy supply parameters and the current cold and heat load parameters include: Feature fusion is performed based on the current cold and heat load parameters and the hourly load prediction values to obtain a predicted cold and heat load parameter sequence in the energy supply future time domain and sorted by time steps; the feature fusion operation takes the current state as the starting point and the correction reference of the prediction trend, so that the prediction sequence generated by the model is more continuous and has actual significance; each row represents a time step and contains fields such as predicted cold load, heat load, and external temperature; The predicted cold and heat load parameter sequence is clustered to obtain a predicted low load cluster, a predicted medium load cluster, and a predicted high load cluster; the minimum value of the predicted cold and heat load parameters in the predicted high load cluster is set as an energy supply determination threshold. The predicted cold and heat load parameter sequence is subjected to cluster analysis to obtain a predicted low load cluster, a predicted medium load cluster and a predicted high load cluster; the input is the hourly predicted cold and heat load sequence output in the previous stage, and the output is the result of dividing these time steps into three cluster labels, each time step being marked as belonging to a low, medium or high load cluster; the data structure is a cluster category label sequence corresponding to the predicted load value, and the method used can be an unsupervised clustering algorithm, such as K-means or density peak clustering, which automatically classifies the predicted values according to distance or density, and the clustering process can be performed separately for cold load and heat load, or a comprehensive load index is used for unified clustering; by converting the continuously changing predicted load sequence into a category determination result, the significant change characteristics of the energy supply load in the time domain are extracted, facilitating subsequent threshold setting and identification of typical load sections.

[0021] If the number of time steps in which the predicted cold and heat load parameter exceeds the energy supply determination threshold in the predicted cold and heat load parameter sequence accounts for more than M% of the total time steps, and there are at least P time periods in which all predicted cold and heat load parameters in each time period exceed the energy supply determination threshold for more than T consecutive time steps, then the energy supply demand in the future time domain is determined to be clear energy supply demand; otherwise, the energy supply demand is determined to be fuzzy energy supply demand. If the number of time steps in which the load value exceeds the above-mentioned energy supply determination threshold in the predicted cold and heat load parameter sequence accounts for more than a specified threshold M% of the total time steps, and at the same time there are at least P time periods that meet the continuity requirement, i.e. the load value in each time period is higher than the threshold for more than T consecutive time steps, then the energy supply future time domain is determined to have clear energy supply demand; otherwise, it is determined to be fuzzy energy supply demand; clear energy supply demand refers to the fact that in a certain time domain in the future, the cooling or heating load presents obvious, high-intensity and persistent characteristics, which exceeds the range of conventional regulation capacity, and the system must clearly dispatch energy storage energy supply equipment to ensure the comfort or production continuity of the user, such demand has strong rigidity and cannot be delayed or reduced; typical characteristics are that high load is concentrated in multiple time periods, the load peak value exceeds a certain threshold and the duration is relatively long, and the corresponding weather conditions, user behavior or calendar type are predictable (such as high temperature at noon in summer, morning and evening peaks in winter, workday work hours); for example, when the meteorological temperature is above 35 degrees, it is a clear cooling demand, and when the meteorological temperature is below 0 degrees, it is a clear heating demand; the specific values of M, P and T are set by professional technicians. Fuzzy energy supply demand refers to the uncertainty, volatility or adjustability of load demand in the future time domain. The system may face low-intensity load demand, or the load exceeds the threshold only for a short time. Flexible adjustment methods such as energy storage and release, load reduction, comfort adjustment, etc. can be used to delay, alleviate or avoid rigid scheduling. For example, during the spring and autumn transition period or during the holiday load uncertainty, the temperature changes greatly during the day and at night. The daytime temperature may reach 25 degrees, but it is not extremely high. The cooling load occasionally exceeds the energy supply threshold, but there is no continuity. The time period above the threshold is less than 20%, and there is no obvious continuous high load section. At this time, some areas have cooling demand during the day, but some areas have no cooling demand all day long, and are in the range of fuzzy energy supply demand.

[0022] Obtain the energy storage and supply equipment performance model and the energy storage tank state data; based on the predicted cooling and heating load parameter sequence and the pre-trained energy saving cost model, the cost analysis is carried out to obtain the direct operation cost, the switching decision cost and the comfort deviation cost; The energy storage and supply equipment performance model is a mathematical model reflecting the efficiency performance of the equipment under different load rates, input power consumption, inlet and outlet water temperature and other operating conditions. It is usually used to predict the energy consumption level, load response capability or energy efficiency coefficient under a specific operating condition. The model is established by historical operation data training or theoretical calculation, which helps the system to accurately calculate the required power consumption and the best operating state of the equipment according to the load demand. The energy storage tank state data refers to the real-time physical data of the current energy storage state of the tank, which is the basis for describing whether the cold storage or heat storage capacity is sufficient and whether it can participate in the next round of energy supply scheduling. The energy storage tank state data includes reference water temperature, average water temperature, total water capacity and other data, which are used to determine whether the energy storage is available, estimate the available load time, control the energy release rate and the energy release cutoff point, and are the key inputs for equipment control logic, energy saving evaluation and switching cost calculation.

[0023] The direct operation cost refers to the electric energy cost consumed by the actual operation of the equipment under the predicted cooling and heating load driving, which is the most basic cost item, reflecting the direct relationship between equipment operation and electricity price. The switching decision cost reflects the energy conversion loss and time delay cost required for the system to switch from the current operating mode (such as relying only on energy storage and supply or direct supply of the main machine) to respond to clear energy supply demand, which reflects the difficulty and cost of the system from the preparation state to the active response state. The comfort deviation cost measures the deviation between the current operating state and the best efficiency point of the equipment, and its economic equivalent value is expressed as: the indirect cost of sacrificing the energy efficiency or user comfort in order to maintain energy supply, which is a kind of energy efficiency penalty expression for non-ideal operation state; The specific steps of cost analysis based on the predicted cooling and heating load parameter sequence and the pre-trained energy saving cost model include: The device predicted power consumption is obtained by calculation based on the predicted cold and heat load parameter sequence and the operation efficiency performance model; the input is a predicted value sequence of the cold and heat load at each time step in the future time domain, and the operation efficiency performance model maps the load and operation temperature to the corresponding power consumption; the output is a predicted value of power corresponding to each time step; it helps to provide a basic energy consumption reference for all scheduling strategies, and facilitates quantification of the total electricity consumption under direct operation or switching operation; The direct operation cost is obtained by calculation based on the device predicted power consumption and the time-of-use electricity price table; The input is the predicted power consumption sequence and the time-of-use electricity price table, and the price table records the corresponding value of the hourly electricity price. The output is the operation cost at each time step and the total operation cost, and the direct operation cost is a sequence of cost values with time labels. The calculation method is point-to-point multiplication, which multiplies the power consumption at a certain time by the corresponding electricity price at the time, and further integrates to obtain the total electricity cost of the entire energy supply time domain; When the clear energy supply demand and the current cold and heat load parameter characteristics are opposite, the switching decision cost is calculated; otherwise, the switching decision cost is set to a standard value; the specific steps for calculating the switching decision cost are: the energy storage tank state data includes the reference water temperature, the average water temperature, and the total water capacity; the energy consumption cost of switching energy is obtained by using physical methods to calculate the energy of the energy storage tank state data; and the total time required to reach the clear energy supply demand is determined based on the pre-trained energy saving cost model, and the switching ability time cost is obtained; the switching decision cost is obtained based on the switching energy internal consumption cost and the switching ability time cost; When the clear energy supply demand and the current cold and heat load parameter characteristics are opposite, it indicates that the system will face a significant change in operating state, for example, the current system is in a cooling state, and the prediction shows that it will enter a clear heating period, or vice versa; such "cooling and heating switching" often accompanies major changes in device operation mode, control strategy, and energy structure; at this time, the switching cost needs to be comprehensively evaluated, and the switching decision cost is calculated, which is essentially the economic loss caused by the physical energy consumption and response time required for the system to switch from the current state to the new state; the standard value is set by humans.

[0024] The switching energy internal consumption cost is calculated based on the energy storage tank state data, the input is the reference water temperature, the average water temperature, and the total water capacity, and the method is to calculate the heat required for water temperature change using thermal physical formula, and the output is the additional required electric power consumption; the switching time required to reach the clear energy supply demand is determined based on the pre-trained energy saving cost model, and the economic cost brought by the delayed energy supply of the system in this period of time is quantified, and the output is the switching time cost; finally, the energy internal consumption cost and the time cost are summed up to obtain the switching decision cost, and the beneficial effect of this step is to evaluate the preparation ability and cost of the system in response to sudden high load scenarios, and to help the scheduling system to predict whether it is worth switching in advance or maintaining the current operating state; The pre-trained energy-saving cost model is used to determine the time and cost required for the system to switch from the current operating state to meet the clear energy supply demand, and its training process is based on historical operation data and scheduling response records; the prototype of the model is a regression type time prediction model, which often uses structures such as gradient boosting tree or long short-term memory network, the input features include the current cooling and heating load level, the state of the energy storage tank (such as average water temperature, temperature difference), the current equipment load rate, the equipment start-up delay time, the environmental meteorological parameters, etc., and the output is the shortest response time (unit: hour or minute) and the corresponding unit energy consumption required to reach the set clear energy supply demand threshold; the training target is to minimize the error between the predicted time and the actual response time, and the mean square error is usually used as the loss function; the training set comes from historical scheduling logs, load prediction records, tank charging and discharging data, and actual energy supply response processes, and needs to cover various typical scenarios such as cooling and heating switching, different seasonal working conditions and equipment combinations; the training termination condition is that the validation set error does not decrease continuously for several rounds or reaches the set minimum error threshold, thereby ensuring the dual stability of the model in generalization ability and response speed, and the model enables the system to predict the switching time, provides quantitative support for scheduling optimization, and avoids blind switching that causes double losses of energy efficiency and experience.

[0025] determine the optimal efficiency operating point and the best load rate of the energy storage energy supply device according to the operating efficiency performance model; calculate the device load rate corresponding to the predicted cooling and heating load parameter sequence; generate an efficiency penalty factor based on the best load rate and the device load rate; calculate based on the efficiency penalty factor and the direct operating cost to obtain the comfort deviation cost; Substitute the predicted cooling and heating load parameter sequence into the operating efficiency performance model to calculate the actual load rate at each time step; generate an efficiency penalty factor based on the deviation between the best load rate and the actual load rate, which reflects the additional energy consumption or comfort loss caused by the device running away from the optimal working condition; multiply the efficiency penalty factor by the direct operating cost at the corresponding time to obtain the comfort deviation cost at that time; associate the device operating efficiency and user comfort to the energy consumption cost, encourage the system to optimize the strategy without sacrificing efficiency and experience, so as to achieve the comprehensive goal of maximizing energy utilization and minimizing operating cost.

[0026] According to the direct operating cost, the switching decision cost and the comfort deviation cost, the rolling optimization is solved; when the energy supply demand is determined as clear energy supply demand, the energy storage energy supply strategy obtained by the rolling optimization is used to control the energy storage energy supply device; when the energy supply demand is determined as fuzzy energy supply demand, the optimal decision selection is made based on the obtained energy storage energy supply strategy and comfort deviation cost to obtain the optimal energy storage energy supply strategy, and the optimal energy storage energy supply strategy is used to control the energy storage energy supply device; When the energy supply demand is determined as clear energy supply demand, a target optimization function is constructed in combination with direct operation cost, switching decision cost and comfort deviation cost; a rolling optimization algorithm is used for solving with the target optimization function as the optimization objective and the energy storage and energy supply equipment performance model and energy storage tank state data as the constraint conditions, to obtain the energy storage and energy supply strategy; When the energy supply demand is determined as clear energy supply demand, the system needs to prioritize load response capability, at this time the target optimization function should be constructed by comprehensively considering direct operation cost, switching decision cost and comfort deviation cost, the input is the predicted cold and heat load parameter sequence, the hourly predicted value of the three cost items, the operation efficiency performance model of the energy storage and energy supply equipment and the current energy storage tank state data, the optimization objective function form is the weighted sum function of the above three cost items, wherein the weight can be dynamically adjusted according to the operation strategy or seasonal adjustment strategy; the constraint conditions include equipment operation boundary (such as upper and lower limits of load rate, start-stop interval), maximum and minimum capacity and water temperature change range of the energy storage tank, and minimum time response required for equipment switching, a rolling optimization algorithm is used for hourly optimization solving of the target function, common methods include model predictive control (MPC) combined with linear or nonlinear programming technology, each optimization outputs the optimal operation strategy for the future number of hours, but only the latest step is executed, and then the prediction and strategy are updated. The beneficial effect of this method is to realize rapid and stable response to sudden high load clear demand, to ensure load safety to meet while reducing overall operation cost, and to realize dynamic collaborative control of energy storage and main engine energy supply.

[0027] When the energy supply demand is determined as a fuzzy energy supply demand, the current energy storage energy supply strategy is obtained; the predicted comprehensive cost target function values generated by executing the current energy storage energy supply strategy and the energy storage energy supply strategy are respectively calculated; the optimal energy storage energy supply strategy is defined based on the predicted comprehensive cost target function values; when the energy supply demand is determined as a fuzzy energy supply demand, the system scheduling enters the flexible response stage, and the difference between the existing operation strategy and the adjustable strategy needs to be fully evaluated to select the cost optimal solution; first, the current energy storage energy supply strategy being executed by the system is obtained, including the current energy storage release curve, the equipment start-stop state and the energy supply allocation mode; based on the preset optimization model and the predicted load, environmental temperature, electricity price and other variables, the target function optimization of the same structure as the clear energy supply determination is executed again to obtain a new recommended energy storage energy supply strategy; then, the comprehensive cost target function values corresponding to the current strategy and the recommended strategy in the entire prediction time domain are respectively calculated, which include three weighted costs of direct operation cost, switching cost (if applicable) and comfort deviation cost; after comparing the target function values of the two strategies, if the cost of the recommended strategy is significantly lower than that of the current strategy, the recommended strategy is defined as the optimal energy storage energy supply strategy, otherwise the original strategy remains unchanged; finally, the selected optimal strategy is used to control the energy storage energy supply equipment, including adjusting the energy storage water pump start-stop, modifying the energy release curve or optimizing the host load rate allocation control instruction, and the like; the beneficial effects of the method are to improve the energy consumption flexible scheduling ability of the system under uncertain load conditions, to realize the balance between economy and flexibility, and to avoid frequent switching or excessive start-stop loss caused by excessive reaction.

[0028] In this embodiment, a certain regional energy supply project is taken as an example. The project relies on the waste heat resources of circulating water to build 10 regional energy stations. Each energy station is equipped with 2 centrifugal water source heat pump units (single unit rated cooling capacity 1000 kW, heating capacity 1100 kW), and is equipped with an underground water storage tank with a capacity of 400 cubic meters for cold storage and heat storage. Historical data of the past 3 consecutive full operation years are called, combined with calendar tags such as holidays and sudden temperature changes, and hourly weather forecasts provided by the meteorological bureau to generate a 48-hour cold load and heat load prediction sequence. Assuming that the average water temperature of the energy storage tank of a certain energy station at the current time is 12°C, the reference cold water outlet temperature is 5°C, and the energy storage state of the tank is 80% cold storage. The chiller is currently operating at partial load (load rate 45%), with an input power of 150 kW and a cooling water inlet temperature of 32°C. According to the prediction data clustering, it is found that there are 10 time periods (each period is continuous ≥3 hours) in the future 48 hours with predicted cold load exceeding 850kw, exceeding the set energy supply threshold (800kW), accounting for 35%. The system determines that this time domain is a clear energy supply demand. Then the system combines the operation efficiency model and the energy saving cost model to calculate the direct operation cost, the switching cost and the comfort deviation cost, and generates an energy storage energy supply strategy based on the rolling optimization algorithm to dispatch the main unit and the tank operation as needed, effectively improving the system energy efficiency, reducing the operation cost and ensuring the energy supply safety and stability. The predicted power consumption is calculated using the operation efficiency performance model, and the main unit refrigeration power consumption is 230kW during the high load period. The direct operation cost is calculated combined with the time-of-use electricity price table. Assuming that the system is in standby state after heat storage and needs to be converted, the switching decision cost is obtained. The equipment will be below 70% load rate in multiple time periods according to the optimization strategy, and the efficiency penalty factor is generated to calculate the comfort deviation cost. Combined with the rolling optimization solution, the following strategy is obtained: during the 0:00~6:00 valley electricity price period, the main unit refrigeration is started at the same time as the water tank cold storage; during the 12:00~18:00 high load and peak electricity price period, the main unit is partially started, combined with the water tank cold release for energy supply; according to the above optimization results, control instructions are issued to the energy storage pump and water source heat pump main unit. If the system determines that it is a fuzzy energy supply demand, the system will calculate the objective function values of the current operation strategy and the above optimization strategy respectively, and if the difference is less than 5%, the current strategy is maintained unchanged to avoid unnecessary switching and improve system stability.

[0029] Embodiment 2, a cooling and heating control system based on energy-saving energy storage, as shown in Figure 1 , comprising: The energy supply data acquisition module comprises a load prediction unit; the load prediction unit is configured to set an energy supply future time domain; energy supply parameters and current cold and heat load parameters of the energy storage energy supply device are acquired before the start of the energy supply future time domain; the energy supply parameters comprise hourly load prediction values, time-of-use electricity price tables and operation efficiency performance models generated based on historical data and weather information; based on the energy supply parameters and the current cold and heat load parameters, prediction is performed to obtain energy supply demand determination and predicted cold and heat load parameter sequences; the energy supply demand determination comprises clear energy supply demand and fuzzy energy supply demand; The energy supply strategy optimization module comprises a cost analysis unit and an energy supply optimization unit; the cost analysis unit is configured to acquire energy storage energy supply device performance models and energy storage tank state data; based on the predicted cold and heat load parameter sequences and a pre-trained energy saving cost model, cost analysis is performed to obtain direct operation cost, switching decision cost and comfort deviation cost; The energy supply optimization unit is configured to perform rolling optimization solution according to the direct operation cost, the switching decision cost and the comfort deviation cost; when the energy supply demand determination is clear energy supply demand, the energy storage energy supply device is controlled by using the energy storage energy supply strategy obtained by the rolling optimization solution; when the energy supply demand determination is fuzzy energy supply demand, optimal decision selection is performed based on the obtained energy storage energy supply strategy and the comfort deviation cost to obtain an optimal energy storage energy supply strategy, and the energy storage energy supply device is controlled by using the optimal energy storage energy supply strategy.

[0030] It should be understood that, for those skilled in the art, improvements or changes can be made according to the above description, and all these improvements and changes shall belong to the protection scope of the appended claims of the present application. The parts not described in detail in the specification belong to the prior art known to those skilled in the art.

Claims

1. A control method for cooling and heating based on energy saving and energy storage, characterized by, The method comprises the following steps: Setting an energy supply future time domain; obtaining energy supply parameters and current cold and heat load parameters of the energy storage energy supply device before the start of the energy supply future time domain; the energy supply parameters include hourly load prediction values generated based on historical data and meteorological information, a time-of-use electricity price table and an operation efficiency performance model; performing prediction based on the energy supply parameters and the current cold and heat load parameters to obtain energy supply demand determination and a predicted cold and heat load parameter sequence; the energy supply demand determination includes clear energy supply demand and fuzzy energy supply demand; Obtaining an energy storage energy supply device performance model and energy storage tank state data; performing cost analysis based on the predicted cold and heat load parameter sequence and a pre-trained energy saving cost model to obtain direct operation cost, switching decision cost and comfort deviation cost; Performing rolling optimization solving according to the direct operation cost, the switching decision cost and the comfort deviation cost; when the energy supply demand determination is clear energy supply demand, controlling the energy storage energy supply device by using the energy storage energy supply strategy obtained by the rolling optimization solving; when the energy supply demand determination is fuzzy energy supply demand, performing optimal decision selection based on the obtained energy storage energy supply strategy and the comfort deviation cost to obtain an optimal energy storage energy supply strategy, and controlling the energy storage energy supply device by using the optimal energy storage energy supply strategy.

2. The energy saving and storage based cooling and heating control method according to claim 1, characterized in that, The specific steps of obtaining the energy supply parameters and the current cold and heat load parameters of the energy storage energy supply device include: Obtaining hourly actual cold and heat load data corresponding to time stamps and a calendar sequence containing special events in at least n complete operation years; obtaining hourly meteorological parameters corresponding to time stamps in the energy supply future time domain; training the actual cold and heat load data and the calendar sequence as training features to obtain a trained cold and heat load prediction model; Inputting the hourly meteorological parameters into the trained cold and heat load prediction model to obtain hourly load prediction values; Obtaining an energy storage energy supply device operation state data set, wherein each piece of data in the operation state data set includes a device load rate, a device input power consumption and a key operation boundary temperature; Constructing an operation efficiency performance model based on the operation state data set and a regression algorithm.

3. The energy saving and energy storage based cooling and heating control method according to claim 2, characterized in that, The specific steps of performing prediction based on the energy supply parameters and the current cold and heat load parameters include: Performing feature fusion based on the current cold and heat load parameters and the hourly load prediction values to obtain a predicted cold and heat load parameter sequence sorted by time steps in the energy supply future time domain; Performing clustering on the predicted cold and heat load parameter sequence to obtain a predicted low load cluster, a predicted medium load cluster and a predicted high load cluster; setting the minimum value of the predicted cold and heat load parameters in the predicted high load cluster as an energy supply determination threshold; If the number of time steps in which the predicted cold and heat load parameters exceed the energy supply determination threshold accounts for more than M% of the total time steps in the predicted cold and heat load parameter sequence, and there are at least P time periods in which all the predicted cold and heat load parameters in each time period continuously exceed the energy supply determination threshold for more than T time steps, the energy supply demand determination of the energy supply future time domain is clear energy supply demand; otherwise, the energy supply demand determination is fuzzy energy supply demand.

4. The energy saving and energy storage based cooling and heating control method according to claim 3, characterized in that, The specific steps of performing cost analysis based on the predicted cold and heat load parameter sequence and the pre-trained energy saving cost model include: The device predicted power consumption is obtained by calculation based on the predicted cold and heat load parameter sequence and the operation efficiency performance model; The direct operation cost is obtained by calculation based on the device predicted power consumption and the time-of-use electricity price table; When the clear energy supply demand and the current cold and heat load parameter characteristics are opposite, the switching decision cost is calculated; otherwise, the switching decision cost is set as a standard value; the specific steps of calculating the switching decision cost are as follows: the reference water temperature, the average water temperature and the total water capacity are contained in the energy storage water tank state data; the switching energy internal loss cost is obtained by energy calculation on the energy storage water tank state data by using a physical method; the total time required to reach the clear energy supply demand is determined based on the pre-trained energy saving cost model, and the switching ability time cost is obtained; the switching decision cost is obtained based on the switching energy internal loss cost and the switching ability time cost; The optimal efficiency operating point and the best load rate of the energy storage energy supply device are determined according to the operation efficiency performance model; the device load rate corresponding to the predicted cold and heat load parameter sequence is calculated; the efficiency penalty factor is generated based on the best load rate and the device load rate; the comfort deviation cost is obtained by calculation based on the efficiency penalty factor and the direct operation cost.

5. The energy saving and energy storage based cooling and heating control method according to claim 4, characterized in that, When the energy supply demand is determined as the clear energy supply demand, the target optimization function is constructed by combining the direct operation cost, the switching decision cost and the comfort deviation cost; the rolling optimization algorithm is used for solving with the target optimization function as the optimization objective and the energy storage energy supply device performance model and the energy storage water tank state data as the constraint conditions, so as to obtain the energy storage energy supply strategy.

6. The energy saving and energy storage based cooling and heating control method according to claim 5, characterized in that, When the energy supply demand is determined as the fuzzy energy supply demand, the current energy storage energy supply strategy is obtained; the predicted comprehensive cost target function values generated by executing the current energy storage energy supply strategy and the energy storage energy supply strategy are calculated respectively; the optimal energy storage energy supply strategy is defined based on the predicted comprehensive cost target function values.

7. A control system for cooling and heating based on energy saving and energy storage, characterized in that, The system is a cold and heat supply control method based on energy saving and energy storage according to any one of claims 1-6, comprising: The energy supply data acquisition module comprises a load prediction unit; the load prediction unit is used for setting an energy supply future time domain; the energy supply parameters and the current cold and heat load parameters of the energy storage energy supply device are acquired before the start of the energy supply future time domain; the energy supply parameters comprise hourly load prediction values generated based on historical data and meteorological information, a time-of-use electricity price table and an operation efficiency performance model; the energy supply demand determination and the predicted cold and heat load parameter sequence are obtained by prediction based on the energy supply parameters and the current cold and heat load parameters; the energy supply demand determination comprises clear energy supply demand and fuzzy energy supply demand; The energy supply strategy optimization module comprises a cost analysis unit and an energy supply optimization unit; the cost analysis unit is used for acquiring the energy storage energy supply device performance model and the energy storage water tank state data; the direct operation cost, the switching decision cost and the comfort deviation cost are obtained by cost analysis based on the predicted cold and heat load parameter sequence and the pre-trained energy saving cost model; When the energy supply demand is determined as the clear energy supply demand, the target optimization function is constructed by combining the direct operation cost, the switching decision cost and the comfort deviation cost; the rolling optimization algorithm is used for solving with the target optimization function as the optimization objective and the energy storage energy supply device performance model and the energy storage water tank state data as the constraint conditions, so as to obtain the energy storage energy supply strategy. When the energy supply demand is determined as the fuzzy energy supply demand, the current energy storage energy supply strategy is obtained; the predicted comprehensive cost target function values generated by executing the current energy storage energy supply strategy and the energy storage energy supply strategy are calculated respectively; the optimal energy storage energy supply strategy is defined based on the predicted comprehensive cost target function values. The energy supply optimization unit is used for solving the rolling optimization according to the direct operation cost, the switching decision cost and the comfort deviation cost, and when the energy supply demand is determined as clear energy supply demand, the energy storage energy supply strategy obtained by the rolling optimization is used to control the energy storage energy supply device; when the energy supply demand is determined as fuzzy energy supply demand, the optimal decision selection is made based on the obtained energy storage energy supply strategy and the comfort deviation cost, the optimal energy storage energy supply strategy is obtained, and the optimal energy storage energy supply strategy is used to control the energy storage energy supply device.

Images