A method for regulating multiple time scales components of rankine-heat pump system supply-demand matching

Abstract

The application discloses a kind of Rankine-heat pump system supply-demand matching multi-time scale component regulation method.The method is first designed according to building load demand Rankine-heat pump system, then the system variable condition mathematical model is established, and the influence law of component concentration on system variable condition performance is analyzed, to determine the optimal matching relationship of cold heat energy and component concentration;On this basis, the correlation of building load influencing factors is analyzed, neural network is trained using historical building load, and a short-term load forecasting model is established, and then according to the weather data, the day-ahead-intra-day load demand of building is predicted.Then, combined with building load demand, the system day-ahead optimization scheduling model is established, to determine the scheduling rule of system day-ahead cold heat energy;According to the day-ahead scheduling rule, the system intra-day layered rolling optimization model is established, to determine the intra-day optimal scheduling rule of cold heat energy.Finally, combined with the optimal component Map of system, the component concentration regulation rule of multi-time scale of system day-ahead-intra-day is determined.

Description

Technical Field

[0001] This invention relates to the field of system performance regulation under varying operating conditions, and in particular to a multi-timescale component control method for supply and demand matching in a Rankine-heat pump system. Background Technology

[0002] The Organic Rankine Cycle (ORC) is a technology that converts low-grade heat energy into electrical energy. Its simple structure and easy operation make it a current hot topic in waste heat power generation technology research. This technology can be widely used in recovering industrial waste heat, solar energy, and geothermal energy. A heat pump is a highly efficient heat transfer device; the heat pump cycle continuously absorbs heat from low-temperature / high-temperature heat sources to provide the building with the required heating / cooling energy. As a novel waste heat utilization device, the Rankine-heat pump system can utilize low-grade heat energy to produce cooling / heating without consuming additional electricity, effectively alleviating grid load and energy shortage problems.

[0003] Because user energy load and ambient temperature fluctuate significantly, Rankine heat pump systems often operate under variable conditions, leading to performance degradation. Taking household air conditioners as an example, when summer outdoor temperatures deviate significantly from design conditions, the desired cooling effect is difficult to achieve, and the system may even malfunction. During variable-condition operation, the system often adjusts condensing pressure and refrigerant mass flow rate, but these methods rarely fundamentally improve the system's performance.

[0004] Non-azeotropic working fluids exhibit temperature glide during phase change heat transfer, which can better match the heat source and cold source, improve the system's thermodynamic performance, and provide operational space for system component control. By adjusting the component concentration of the Rankine-type heat pump system, on the one hand, the capacity of the heat pump system can be adjusted to match its output of cooling / heating energy with user needs; on the other hand, the compatibility of the ORC system with the environment can be improved, maximizing its annual electrical energy output and enhancing the overall system performance.

[0005] However, the demand for cooling, heating, and electricity in buildings fluctuates significantly, and the predicted values ​​of these loads often deviate from real-time energy consumption. Furthermore, the demand for cooling / heating energy has a large time constant and a slow response time, while electrical energy responds very quickly. This greatly increases the difficulty of system component regulation. Therefore, this paper proposes a multi-timescale component regulation method that combines short-term building load forecasting with multi-timescale rolling optimization to achieve optimized energy scheduling at multiple response time levels in the Rankine-heat pump system and improve the overall system performance. Summary of the Invention

[0006] To address the aforementioned issues in the performance regulation of Rankine heat pump systems under varying operating conditions, this invention provides a multi-timescale component control method for supply and demand matching in Rankine heat pump systems. This method adjusts the working fluid components of the system at multiple timescales to achieve efficient matching between system energy supply and building energy consumption, thereby improving the system's performance under varying operating conditions.

[0007] The specific implementation steps of the adopted technical solution are as follows:

[0008] (1) Record the historical cooling, heating and power load demand data of the building, and design the Rankine-heat pump system accordingly, and determine the main structural parameters and equipment capacity of the system's steam generator, expander, condenser, working fluid pump, compressor, evaporator, thermal expansion valve and component regulating device.

[0009] (2) Based on the system design results, establish the variable operating condition mathematical model of the main components, and based on the heat and mass transfer relationship between the components, establish the variable operating condition mathematical model of the Rankine-heat pump system.

[0010] (3) Analyze the coupling effect of component concentration and ambient temperature on the system's performance under varying operating conditions. With the goal of achieving the optimal overall performance of the Rankine-heat pump system, determine the optimal components under different load outputs and ambient temperatures, and obtain a Map of the optimal components.

[0011] (4) Analyze the factors that have a significant impact on the building’s cooling, heating and electricity load, establish a neural network prediction model for the building’s short-term load, and use the building’s historical cooling, heating and electricity load to train the neural network and determine the optimal neural network prediction model.

[0012] (5) Based on the real-time weather forecast data from the meteorological station, the short-term load forecasting model and the forecasting error model are used to forecast the short-term load demand for cooling, heating and electricity before and during the day of the building, with the time interval Δt as the model forecasting time step.

[0013] (6) Combining the variable operating condition mathematical model of the Rankine-heat pump system, and taking the day-ahead cooling, heating and power load demand of the building as the constraint, and taking the system energy utilization efficiency and investment payback period as the objectives, a day-ahead optimization scheduling model of the system is established. The day-ahead optimization scheduling model is then optimized to determine the scheduling plan of the day-ahead cooling, heating and power load of the system.

[0014]

[0015] The constraints are as follows:

[0016]

[0017] (7) Based on the daily scheduling plan, with the goal of minimizing system revenue fluctuations, an intraday hierarchical multi-timescale rolling optimization scheduling model is established. The intraday hierarchical scheduling plan for the system's cold, hot and electrical energy is determined through rolling optimization. In the intraday hierarchical optimization scheduling model, the upper layer mainly regulates cold and hot energy with longer response times. On the basis of meeting the cold and hot energy demand of the upper layer, the lower layer mainly regulates electrical energy with shorter response times to reduce system fluctuations.

[0018] The upper-level optimization objective is:

[0019]

[0020] The lower-level optimization objective is:

[0021]

[0022] (8) Based on the day-to-day scheduling pattern of the system's cooling, heating and electrical energy, and combined with the Map of the system's cooling, heating energy and optimal components, determine the multi-timescale control scheme of the system components that meet the building's load requirements by day-to-day layering.

[0023] In step (1), the Rankine heat pump system is designed according to the heat-driven power consumption model. Based on the building's historical cooling, heating, and power load requirements, the building's cooling and heating loads are met first, and any excess / insufficient power is output to / provided by the power grid.

[0024] Furthermore, in step (1), the designed Rankine-heat pump system has a component adjustment device and can adjust the component concentration of the non-azeotropic working fluid in real time according to the energy load demand of the building, thereby changing the output of cooling, heating and electrical energy, so that the system's energy supply matches the building's load demand.

[0025] In step (2), mathematical models of each component should be established first. Static models can be used to establish the working fluid pump, expander, compressor, and thermostatic expansion valve; dynamic models can be established for the evaporator and condenser. Then, a system model is established based on the physical coupling relationship between each component.

[0026] In step (3), the comprehensive performance of the Rankine-heat pump system refers to taking energy utilization efficiency and investment payback period as the thermal economic performance indicators of the Rankine-heat pump system, then using the expert method to determine the weighting factors of the two indicators, and finally taking the weighted value of the two as the comprehensive performance of the system.

[0027] Furthermore, the overall performance of the Rankine-heat pump system is used as the criterion for judgment. Based on this, the optimal components of the system under different cooling and heating energy and ambient temperatures are determined, and each optimal component is integrated into a Map.

[0028] In step (4), based on the building's historical cooling, heating, and electricity load parameters, the Spearman correlation coefficient method is used to analyze the correlation between outdoor, indoor, and time-related parameters and the cooling, heating, and electricity loads. Parameters with high load correlation are selected and used as the input layer of the neural network in the short-term load prediction model for training. Then, a genetic algorithm is used to optimize the weights and thresholds of the neural network to improve the prediction accuracy of short-term loads, forming a neural network prediction model with better accuracy.

[0029] In step (5), the indoor summer and winter temperatures are set to 26°C and 18°C, respectively.

[0030] When forecasting the day-ahead load demand, since temperature changes are generally continuous, the cooling, heating, and electricity demand of buildings typically fluctuates within a certain range. Therefore, the previous day's meteorological data and building cooling, heating, and electricity load data provide some guidance for forecasting the current day's load. Specifically, the day-ahead forecast combines the meteorological station's day-ahead weather forecast data with a short-term building load neural network model to obtain the building's day-ahead load forecast, with a forecast interval of 1 hour, thus obtaining the hourly day-ahead cooling, heating, and electricity load forecast.

[0031] Furthermore, when predicting intraday load demand, short-term forecasts are made by combining real-time data of the day. On the one hand, the daily weather forecast data is used to predict cooling, heating, and electricity load demand; on the other hand, normal prediction error is added to reflect prediction deviation. The time interval for intraday cooling and heating load forecasts is 15 minutes, and the time interval for intraday electricity forecasts is 5 minutes.

[0032] In step (6), a day-ahead optimization scheduling model for the system is established, wherein the optimization scheduling objective is... For energy efficiency; W net (t) represents the net output power; Q cool (t) represents the cooling capacity; Q heat (t) represents the heat output; Q in (t) represents the input heat power; PBP represents the payback period; k represents the interest rate; C tot C represents the total investment cost; w For electricity price; C cool Price per unit of cooling capacity; C heat The unit heating cost is LT; the system lifespan is LT; the time interval described in step (5) is Δt; f k This is the cost coefficient.

[0033] In the set constraints, W net (t) represents the net output power; W grid (t) represents the power transmitted through the power grid; W bp (t) represents the user's required electrical power. cool(t) represents the cooling capacity; Q bc (t) represents the user's cooling load; Q heat (t) represents the heat output; Q bh (t) represents the user's heat load.

[0034] Furthermore, when establishing the day-ahead optimization scheduling model for the Rankine-heat pump system, the day-ahead scheduling cycle is 24 hours with 1-hour intervals. The day-ahead optimization determines the scheduling plan for the system's cooling, heating, and electrical energy for the next working day.

[0035] In step (7), according to the intraday hierarchical multi-timescale rolling optimization scheduling model, the rolling optimization continuously repeats the sampling-optimization, resampling-reoptimization process as time progresses. The intraday cold and heat energy scheduling cycle is 1 hour with 15-minute intervals, and the power energy scheduling cycle is 15 minutes with 5-minute intervals to achieve coordinated scheduling of cold and heat / power energy at different response time levels. The scheduling rules of cold, heat, and power energy in each intraday scheduling cycle are determined through rolling optimization.

[0036] In step (8), based on the day-to-day multi-timescale scheduling pattern of the system's cooling, heating, and electrical energy, when querying the Map of the Rankine-heat pump system's cooling, heating, and electrical energy and the optimal working fluid composition, a two-dimensional difference function is used to determine the component concentration corresponding to the day-to-day optimal scheduling plan of cooling, heating, and electrical energy.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] (1) Considering the large fluctuations in building energy load and the poor performance of traditional Rankine-heat pump systems under varying operating conditions, this invention introduces the concept of component regulation. The component concentration of the Rankine-heat pump system is adjusted in real time according to the building load demand and ambient temperature, thereby improving the operating efficiency of the Rankine-heat pump system while achieving adaptive supply and demand matching.

[0039] (2) The present invention constructs a multi-timescale component control method from day to day, and continuously adjusts the working fluid components of the system through rolling optimization, thereby improving the accuracy and timeliness of system component control.

[0040] (3) The present invention takes into account the difference in response time between cold and hot energy and electrical energy. It adopts a longer time interval for cold and hot energy with large thermal inertia and slow response speed, and a shorter time interval for electrical energy with fast response speed, so as to realize the coordinated optimization scheduling of energy with different response time levels and make the system supply and demand matching control more precise. Attached Figure Description

[0041] Figure 1 This is a flowchart of a multi-timescale component control method for supply and demand matching in a Rankine-heat pump system according to the present invention.

[0042] Figure 2 This is a schematic diagram of the design of the component-adjustable Rankine heat pump system in this invention.

[0043] Figure 3 This is the method flow for implementing short-term load forecasting in the embodiments of the present invention.

[0044] Figure 4 This is a schematic diagram of the actual daily cooling, heating and electrical load of a building in an embodiment of the present invention.

[0045] Figure 5 This is a schematic diagram of the day-ahead and intraday optimized scheduling strategies in an embodiment of the present invention. Detailed Implementation

[0046] The following detailed description, in conjunction with schematic diagrams, illustrates a multi-timescale component control method for supply and demand matching in a Rankine-heat pump system, which represents a preferred embodiment of the invention. It should be understood that those skilled in the art can modify the invention described herein while still achieving its advantageous effects. Therefore, the following description should be construed as an explanation for those skilled in the art and not as a limitation thereof.

[0047] Figure 1 This is a flowchart of the overall method of the present invention.

[0048] A multi-timescale component control method for supply and demand matching in a Rankine heat pump system is disclosed. This Rankine heat pump system incorporates multiple energy inputs (low-to-medium grade heat energy such as geothermal energy, and electrical energy), and utilizes a Rankine cycle to generate electricity to drive the heat pump system, thereby providing various forms of energy output (cooling, heating, and electricity) to the building. By controlling the system's components at multiple timescales, including day-ahead and intraday periods, energy supply and demand matching is achieved. The building's energy load involves both electrical and cooling / heating loads. This multi-timescale component control method for supply and demand matching in a Rankine heat pump system includes the following steps:

[0049] Step (1): Obtain historical cooling, heating and power load demand data for the building, and design the Rankine-heat pump system based on the historical load demand data.

[0050] In a Rankine heat pump system, the main components include an expander, condenser, working fluid pump, compressor, evaporator, thermostatic expansion valve, and component regulating device. The main structural parameters and equipment capacity of each component of the Rankine heat pump system need to be calculated through design.

[0051] The designed Rankine heat pump system has a component adjustment device that can adjust the component concentration of the non-azeotropic working fluid in real time according to the building's energy load demand, thereby changing the output of cooling, heating and electrical energy, so that the system's energy supply matches the building's load demand. Figure 2This is a schematic diagram of the component-adjustable Rankine heat pump system design in this invention, illustrating the working process and basic principle of the Rankine heat pump system.

[0052] Step (2): Based on the system design results, establish the variable operating condition mathematical model of the main components, and establish the variable operating condition mathematical model of the Rankine-heat pump system based on the heat and mass transfer relationship between the components. The system model is established through the physical coupling relationship of each component.

[0053] Step (3): Analyze the coupling effect of component concentration and ambient temperature on the system's performance under varying operating conditions. With the goal of achieving the optimal overall performance of the Rankine-heat pump system, determine the optimal components under different load outputs and ambient temperatures, and obtain the optimal component map.

[0054] Step (4): Analyze the factors that have a significant impact on the building’s cooling, heating and electricity loads, establish a neural network prediction model for the building’s short-term loads, and train the neural network using the building’s historical cooling, heating and electricity loads to determine the optimal neural network prediction model.

[0055] Based on the building's historical cooling, heating, and electricity load parameters, the Spearman correlation coefficient method was used to analyze the correlation between outdoor, indoor, and time parameters and cooling, heating, and electricity loads. Parameters with high load correlation were selected and used as the input layer of the neural network in the short-term load forecasting model for training.

[0056] By using a genetic algorithm to optimize the weights and thresholds of a neural network, the prediction accuracy of short-term loads is improved, resulting in a neural network prediction model with better accuracy.

[0057] Step (5): Preset the indoor summer and winter temperatures to 26℃ and 18℃ respectively.

[0058] like Figure 3 This paper demonstrates the methodology for short-term load forecasting. Short-term load forecasting can be divided into day-ahead load demand forecasting and intraday load demand forecasting at different time scales. The data sources for obtaining demand forecasts are both related and different.

[0059] When forecasting the day-ahead load demand, since temperature changes are generally continuous, the cooling, heating, and electricity demand of buildings typically fluctuates within a certain range. Therefore, the previous day's meteorological data and building cooling, heating, and electricity load data provide some guidance for forecasting the current day's load. Specifically, the day-ahead forecast combines the meteorological station's day-ahead weather forecast data with a short-term building load neural network model to obtain the building's day-ahead load forecast, with a forecast time interval of 1 hour, i.e., hourly day-ahead cooling, heating, and electricity load forecast.

[0060] When forecasting intraday load demand, short-term forecasts are made by combining real-time data of the day. First, the day's weather forecast data is acquired to predict cooling, heating, and electricity load demand. Simultaneously, a normality error is introduced to reduce forecast deviation. The time interval for intraday cooling and heating load forecasts is 15 minutes, and the time interval for intraday electricity load forecasts is 5 minutes.

[0061] The above process yields an optimized short-term energy consumption forecast, which is close to the actual daily cooling, heating, and electrical load of the building. For example... Figure 4 The diagram shows the actual daily heating, cooling, and electrical load of a building. The heating load fluctuates dramatically over time, with peak load occurring from 6:00 AM to 9:00 PM, and the load dropping to zero during the off-peak hours. The building's electrical load also exhibits a clear peak-valley characteristic, with peak electricity consumption from 9:00 AM to 6:00 PM, accompanied by fluctuations, and off-peak periods during the remaining time.

[0062] Step (6): Establish a day-ahead optimization scheduling model for the system.

[0063] The optimization scheduling objective is:

[0064]

[0065] In the optimized scheduling objective For energy efficiency; W net (t) represents the net output power; Q cool (t) represents the cooling capacity; Q heat (t) represents the heat output; Q in (t) represents the input heat power; PBP represents the payback period; k represents the interest rate; C tot C represents the total investment cost; w For electricity price; C cool Price per unit of cooling capacity; C heat The unit heating cost is LT; the system lifespan is LT; the time interval described in step (5) is Δt; f k This is the cost coefficient.

[0066] The constraints are as follows:

[0067]

[0068] In the set constraints, W net (t) represents the net output power; W grid (t) represents the power transmitted through the power grid; W bp (t) represents the user's required electrical power. cool (t) represents the cooling capacity; Q bc (t) represents the user's cooling load; Q heat (t) represents the heat output; Q bh (t) represents the user's heat load.

[0069] When establishing the day-ahead optimization scheduling model for the Rankine-heat pump system, the day-ahead scheduling cycle is 24 hours, with data updated on a rolling basis at 1-hour intervals. A day-ahead scheduling plan for cooling, heating, and power is formulated with 24 time intervals as one cycle. The day-ahead optimization determines the scheduling plan for the system's cooling, heating, and power energy for the next working day.

[0070] Step (7): According to the intraday hierarchical multi-timescale rolling optimization scheduling model, the rolling optimization continuously repeats the process of sampling-optimization, resampling-re-optimization as time progresses.

[0071] Figure 5 The diagram shows the day-ahead and intraday optimization scheduling strategies in the rolling optimization scheduling model.

[0072] Because the response times for regulating heating and cooling energy and electrical energy are different, different time intervals are adopted in the daytime forecasting and intraday scheduling of heating and cooling energy and electrical energy loads, respectively.

[0073] For hot and cold energy with high thermal inertia and slow response speed, a longer scheduling cycle and time interval are adopted. Data is updated on a rolling basis at 15-minute intervals. The daily hot and cold energy scheduling cycle is taken as 1 hour, and 4 time intervals are taken as one cycle. That is, a daily optimization strategy for hot and cold energy is formulated every 1 hour.

[0074] For fast-responding power, a shorter time interval is used, with data updated on a rolling basis at 5-minute intervals. The power scheduling cycle is taken as 15 minutes, and 3 time intervals are taken as one cycle, that is, a daily power optimization strategy is formulated every 15 minutes.

[0075] This enables coordinated scheduling of heating / cooling and electrical energy at different response times. Rolling optimization is used to determine the scheduling patterns of heating, cooling, and electrical energy within each scheduling cycle throughout the day.

[0076] According to the requirements of the intraday hierarchical multi-timescale rolling optimization scheduling model...

[0077] The upper-level optimization objective is:

[0078]

[0079] In the aforementioned upper-level optimization objective, F1 represents the unit cost of cooling and heating energy; t0 is the starting time, d is the number of time intervals, and Δt1 is the intraday cooling and heating energy scheduling interval, taken as Δt1 = 5 min; C w For electricity price; W grid (t) represents the power transmitted through the power grid; μ cool The penalty price is adjusted based on the amount of cooling capacity, ΔQ cool (t1) represents the cooling capacity adjustment at time t1; μ heat Adjusting the penalty price for heat, ΔQheat (t1) represents the amount of heat adjustment at time t1.

[0080] The lower-level optimization objective is:

[0081]

[0082] In the lower-level optimization objective, F2 represents the unit energy cost; similarly, with time t0 as the starting time, n as the time interval, and Δt2 as the daily energy dispatch interval, Δt2 = 15 min is taken; C w For electricity price; W grid (t) represents the power transmitted through the power grid; μ cool The penalty price is adjusted based on the amount of cooling capacity, ΔQ cool (t2) represents the cooling capacity adjustment at time t2; μ heat Adjusting the penalty price for heat, ΔQ heat (t2) represents the amount of heat adjustment at time t2.

[0083] Based on the above, a daytime hierarchical multi-timescale rolling optimization scheduling model is obtained. By combining the system's daytime hierarchical multi-timescale optimization scheduling model with the system's cold and hot energy and optimal working fluid composition Map described in step (3), the daytime-intraday multi-timescale scheduling law of the system's cold, hot, and electrical energy can be obtained.

[0084] Step (8): Based on the day-to-day multi-timescale scheduling patterns of the system's heating, cooling, and electrical energy, when querying the Map of the Rankine-heat pump system's heating, cooling, and electrical energy and the optimal working fluid composition, a two-dimensional difference function is used to determine the component concentration corresponding to the day-to-day optimal scheduling plan for heating, cooling, and electrical energy. The component concentration regulation patterns of the system at multiple timescales from day to day are summarized.

[0085] The above are merely illustrative embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and its improved concept, should be covered within the scope of protection of the present invention.

Claims

1. A multi-time scale component regulation method for supply-demand matching of a Rankine-heat pump system, characterized in that, It comprises the following steps: (1) According to the historical cooling, heating and power load demand of the building, the Rankine-heat pump system is designed to determine the main structural parameters and equipment capacity of the system components such as steam generator, expander, condenser, working fluid pump, compressor, evaporator, thermal expansion valve and component adjusting device; (2) According to the system design results, the variable condition mathematical model of the main components is established, and the variable condition mathematical model of the Rankine-heat pump system is established according to the heat and mass transfer relationship between the components; (3) The coupling influence law of component concentration and environmental temperature on the variable condition performance of the system is analyzed, the optimal component of the system under different load output and environmental temperature is determined to obtain the Map graph of the optimal component with the optimal comprehensive performance of the Rankine-heat pump system as the target; (4) The factors that have greater influence on the building cooling, heating and power load are analyzed, the neural network prediction model of the short-term load of the building is established, and the optimal neural network prediction model is determined by training the neural network with the historical cooling, heating and power load of the building; (5) According to the real-time weather forecast data of the weather station, the short-term load prediction model is used to predict the building's cooling, heating and power load demand in advance and within a day with the time interval Δt as the model prediction time step; (6) The variable condition mathematical model of the Rankine-heat pump system is combined, the system energy utilization efficiency and investment recovery period are taken as the target, and the system day-ahead optimal scheduling model is established with the building's cooling, heating and power load demand in advance as the constraint condition, so that the day-ahead cooling, heating and power load scheduling plan of the system is determined by optimizing the day-ahead optimal scheduling model; The optimization scheduling target is: The set constraint condition is: The optimization scheduling target, η is energy utilization rate; W net (t) is net output power; Q cool (t) is refrigeration capacity; Q heat (t) is heating capacity; Q in (t) is input heat power; PBP is payback period; k is interest rate; C tot is total investment cost; C w is electricity price; C cool is unit refrigeration capacity price; C heat is unit heating capacity price; LT is system life; Δt is time interval in step (5); f k is cost coefficient; W net (t) is the net output power; W grid (t) is the grid transmission power; W bp (t) is the user demand electric power; Q cool (t) is the refrigeration capacity; Q bc (t) is the user cold load; Q heat (t) is the heating capacity; Q bh (t) is the user heat load; (7) On the basis of following the day-ahead scheduling plan, the day-ahead rolling optimal scheduling model is established with the minimum system revenue fluctuation as the target, and the day-ahead hierarchical scheduling plan of the cooling, heating and power energy of the system is determined by rolling optimization; In the day-ahead hierarchical optimal scheduling model, the upper layer mainly regulates and controls the cooling and heating energy with a longer response time, and the lower layer mainly regulates and controls the power energy with a shorter response time to reduce the fluctuation of the system; The upper layer optimization target is: In the upper layer optimization objective, F1 is the unit cold and heat energy cost; t0 is the starting time, d is the number of time intervals, Δt1 is the intra-day cold and heat energy scheduling interval, and Δt1=5 min is taken; w W is the electricity price; grid (t) is the grid transmission power; μ cool is the cold adjustment penalty price, ΔQ cool (t1) is the cold adjustment amount at t1; μ heat is the heat adjustment penalty price, ΔQ heat (t1) is the heat adjustment amount at t1; The lower layer optimization target is: In the lower layer optimization objective, F2 is the unit cost of electricity; similarly, t0 is taken as the starting time, n is the number of time intervals, Δt2 is the interval of intraday electricity scheduling, and Δt2=15 min; w C is the electricity price; W grid (t) is the grid transmission power; μ cool is the cold adjustment penalty price, ΔQ cool (t2) is the cold adjustment amount at t2; μ heat is the heat adjustment penalty price, ΔQ heat (t2) is the heat adjustment amount at t2; (8) According to the day-ahead-in-day scheduling law of the cooling, heating and power energy of the system, and combining the Map graph of the cooling, heating and power energy and the optimal component, the day-ahead-in-day hierarchical multi-time scale regulation scheme of the system component that meets the building load demand is determined.

2. The method of claim 1, wherein the method is a Rankine-heat pump system supply-demand matching multi-time scale component regulation method. The Rankine-heat pump system operates in the mode of heat determining power, and the excess / deficient power is outputted / provided to the power grid on the basis of meeting the building cooling and heating load; At the same time, the system adjusts the component concentration of the non-azeotropic working fluid in real time according to the energy load demand of the building, changes the output of the cooling, heating and power energy, and matches the energy supply of the system with the load demand of the building.

3. The method of claim 1, wherein the method is a Rankine-heat pump system supply-demand matching multi-time scale component regulation method. In step (3), the energy utilization efficiency and net present value are taken as the thermal economic performance indicators of the Rankine-heat pump system, the weight factors of the two indicators are determined by expert method, the weighted values of the two indicators are taken as the comprehensive performance of the system, and the system optimal component under different cooling, heating and power energy and environmental temperature is determined as the judgment basis.

4. The method of claim 1, wherein: In the step (4), according to the building historical load parameters, the correlation of outdoor, indoor, time and other parameters with the cold, heat and power load is analyzed by using the Spearman correlation coefficient method, the parameters with high correlation degree are taken as the input layer of the neural network in the short-term load prediction model, and the genetic algorithm is used to optimize the weight and threshold of the neural network, so as to improve the prediction accuracy of the short-term load.

5. The method of claim 1, wherein: In the step (5), the indoor temperature in summer and winter is set to 26 DEG C and 18 DEG C respectively, the short-term load neural network model of the building and the weather forecast data of the weather station are used to predict the cold, heat and power load of the building in the day, and the time interval of the prediction is 1 hour; when the load demand in the day is predicted, on one hand, the cold, heat and power load demand is predicted by using the weather forecast data of the day, on the other hand, the normal error is added to reflect the prediction deviation, in addition, the time interval of the short-term cold and heat load prediction is 15 minutes, and the time interval of the power prediction is 5 minutes.

6. The Rankine-heat pump system supply-demand matching multi-time scale component regulation method according to claim 1, characterized in that: In the step (6), when the day-ahead optimization scheduling model of the Rankine-heat pump system is established, the day-ahead scheduling cycle is 24 hours and the interval is 1 hour, and the scheduling plan of the day-ahead cold, heat and power of the next working day of the system is determined by optimization.

7. The Rankine-heat pump system supply-demand matching multi-time scale component regulation method according to claim 1, characterized in that: In the step (7), the short-term cold and heat energy scheduling cycle is 1 hour and the interval is 15 minutes, the power scheduling cycle is 15 minutes and the interval is 5 minutes, so as to realize the collaborative scheduling of the cold, heat and power with different response time levels, and the scheduling rule of the cold, heat and power in each scheduling cycle in the day is determined by rolling optimization.

8. The Rankine-heat pump system supply-demand matching multi-time scale component regulation method according to claim 1, characterized in that: In the step (8), when the Rankine-heat pump system cold and heat energy and the optimal working medium component Map chart are inquired, the two-dimensional difference function is used to determine the component concentration corresponding to the optimal scheduling plan of the day-ahead and short-term cold, heat and power.

Images