Capacity configuration design method for medium-deep layer buried pipe heat pump system

Abstract

The invention discloses a medium-deep buried pipe heat pump system capacity configuration design method, which comprises the following steps of: acquiring short-term heat exchange test data and heat exchange theoretical data of a medium-deep buried pipe, and performing data inversion to obtain comprehensive thermophysical parameters of a rock-soil body; building a building heat load prediction model according to the historical heat load, the historical state data and the building information to obtain a predicted heat load, determining a heat supply mode according to the predicted heat load, and performing dynamic heat supply simulation by adopting a long-term heat exchange model of the medium-deep buried pipe heat pump system to obtain configuration operation data, and optimizing the medium-deep layer buried pipe heat pump system long-term heat exchange model according to the configuration operation data, and outputting project system capacity configuration by the optimized medium-deep layer buried pipe heat pump system long-term heat exchange model. According to the method, the precision and efficiency of capacity configuration of the medium-deep buried pipe heat pump system can be improved, meanwhile, good interpretability is achieved, and the method has great significance in reducing the construction cost, saving energy and reducing emission.

Description

Technical Field

[0001] The present invention relates to the field of geothermal energy technology, and in particular to a capacity configuration design method for a medium-deep buried pipe heat pump system. Background Art

[0002] my country has rich and diverse geothermal resources, among which medium and deep geothermal energy reserves are abundant and have great potential for development and utilization. Medium and deep geothermal energy utilization technology has become an emerging geothermal heating technology due to its clean, efficient and sustainable characteristics. It is of great significance to solve the atmospheric pollution problem of winter heating in northern regions and the combined supply of cold and heat in southern regions.

[0003] Since the heat exchange performance of medium-deep underground pipes is greatly affected by the thermal physical parameters of underground rock and soil, the traditional heat exchange model mainly relies on the empirical valuation of the thermal physical parameters of rock and soil, which deviates from the actual heat exchange and heating power; at the same time, the building heat load, as the main basis for the capacity configuration of the heat pump system, is affected by a variety of building and environmental factors. The existing capacity configuration design method is mainly based on the static building heat load, which is difficult to accurately reflect the dynamic demand in actual operation, thereby reducing the system efficiency and increasing the operating cost; in addition, the existing design methods mostly focus on static capacity configuration, lack of comprehensive consideration of the dynamic operation characteristics of the system, resulting in low actual operating efficiency. The present invention proposes a capacity configuration design method for medium-deep underground pipe heat pump systems. By combining short-term heat exchange test data, theoretical data and building heat load prediction models, the thermal physical parameters of rock and soil are accurately obtained, and the system capacity configuration is optimized based on dynamic heat load simulation, overcoming the shortcomings of existing optimization and adjustment methods, thereby improving the system operation efficiency and reducing the operating cost. The method has broad application prospects and can provide scientific basis and technical support for the design and operation of medium-deep underground pipe heat pump systems. Summary of the invention

[0004] The purpose of the present invention is to provide a capacity configuration design method for a medium-deep buried pipe heat pump system.

[0005] To achieve the above object, the present invention is implemented according to the following technical solutions:

[0006] The present invention comprises the following steps:

[0007] Obtain short-term heat exchange test data and heat exchange theoretical data of medium-deep buried pipes, and perform data inversion based on the heat exchange test data and the heat exchange theoretical data to obtain comprehensive thermal physical property parameters of rock and soil;

[0008] Obtain historical heat load and historical status data of similar buildings, construct a building heat load prediction model based on the historical heat load, the historical status data and building information, and input the building status information to be predicted into the building heat load prediction model to obtain a predicted heat load;

[0009] Determine the heating mode according to the hourly dynamic information of the predicted heat load, and use the long-term heat exchange model of the medium-deep buried pipe heat pump system to perform dynamic heating simulation to obtain configuration operation data; the configuration operation data includes system capacity configuration and operation temperature data;

[0010] The long-term heat exchange model of the medium-deep buried pipe heat pump system is optimized according to the configuration operation data, and the project system information to be configured is input into the optimized long-term heat exchange model of the medium-deep buried pipe heat pump system to output the project system capacity configuration.

[0011] Furthermore, the method for obtaining the comprehensive thermophysical property parameters of the rock and soil mass includes:

[0012] Construct medium-deep buried pipe test holes at the project site, and conduct short-term heat exchange tests based on the medium-deep buried pipe test holes to obtain heat exchange test data and original temperature distribution data of the project site;

[0013] According to the original temperature distribution of the project site and the parameters of the buried pipe test hole, a short-term heat exchange simulation of the medium-deep buried pipe was carried out to obtain heat exchange simulation data;

[0014] The heat exchange test data and heat exchange simulation data are back-calculated and fitted, and the comprehensive thermal physical property parameters of the rock and soil are adjusted according to the back-calculation and fitting results. The adjusted comprehensive thermal physical property parameters of the rock and soil are input into the short-term heat exchange simulation model of the medium-deep buried pipe to obtain the adjusted heat exchange simulation data;

[0015] Repeat the above back-calculation fitting and parameter adjustment operations until the back-calculation objective function is minimized, and output the corresponding comprehensive thermal physical properties parameters of the rock and soil mass.

[0016] Furthermore, the method for obtaining the predicted heat load includes:

[0017] Obtain historical heat load and historical status data of similar buildings, input the historical status data into the state function to obtain the heating state factor, and construct the heating structure factor according to the building information;

[0018] The historical heat load, historical status data, heating status factor and heating structure factor are combined to obtain a comprehensive prediction set, and the comprehensive prediction set is divided into a training set and a test set according to a ratio of 7:3;

[0019] Construct a building heat load prediction model, which includes genetic algorithm, BP neural network, loss function and grid search algorithm;

[0020] The genetic algorithm optimizes the initial weights and biases of the BP neural network; the BP neural network learns the highly nonlinear relationship between historical heat load, historical state data, heating state factor and heating structure factor, and predicts the building heat load based on the input state data, heating state factor and heating structure factor; the grid search algorithm tunes the hyperparameters of the building heat load prediction model; the loss function measures the difference between the predicted value and the true value, and the expression is:

[0021]

[0022] Among them, α and β are the weights of the loss function, δ is the error hyperparameter used to control the error threshold, and Q i is the true value of heat load in stage i, is the predicted value of heat load in stage i, and m is the number of samples

[0023] The state information of the building to be predicted is input into the building heat load prediction model to obtain the predicted heat load.

[0024] Furthermore, the method for obtaining the comprehensive prediction set includes:

[0025] Input the historical state data into the state function to obtain the heating state factor, the expression is:

[0026]

[0027] Among them I sta，i is the heating state factor of stage i, u1 is the temperature weight, u2 is the level weight, T ei is the outdoor ambient temperature in stage i, T hi is the indoor temperature of the historical building in stage i, T ni is the indoor temperature of the proposed building in phase i, H0 is the standard humidity, H hi is the indoor humidity of the historical building in stage i, H ni is the indoor humidity of the proposed building in phase i, is the heating level of the proposed building, Heating rating for historic buildings;

[0028] The heating structure factor is constructed based on the building information, and the expression is:

[0029]

[0030] Among them I str is the heating structure factor, u3 is the space weight, u4 is the insulation weight, A n is the planned building area, V n is the volume of the proposed building space, High n A is the proposed building height, h is the historical building area, Vh For the volume of historical building space, High h For the historical building, is the insulation grade of the proposed building, R n is the air renewal rate of the proposed building, is the insulation level of historical buildings, R h Air renewal rate for historic buildings;

[0031] The historical heat load, historical status data, heating status factors and heating structure factors are combined to obtain a comprehensive prediction set.

[0032] Furthermore, the method for determining the heating mode includes:

[0033] Obtain hourly dynamic information of predicted heat load, define heat load change greater than 5% of design heat load as a heat load update, calculate heat load update dynamic index, set heat load update dynamic index threshold, compare heat load update dynamic index with corresponding heat load update dynamic index threshold to determine heating mode;

[0034] The method for determining the heating mode is specifically as follows: dynamic heating is performed when more than 1 / 2 of the heat load dynamic indicators exceed the corresponding heat load dynamic indicator threshold, otherwise constant heating is performed; the specific plan for dynamic heating is that the medium-deep buried pipe heat pump system bears all the dynamic heat loads of the building; the specific plan for constant heating is that the medium-deep buried pipe heat pump system bears the constant heat load of the building foundation, and the remaining heat load is borne by the auxiliary heat source; the constant heat load is determined according to the optimal heating efficiency of the medium-deep buried pipes.

[0035] Furthermore, the method for obtaining the configuration operation data includes:

[0036] Determine the building design heat load and the annual cumulative heat load based on the hourly dynamic information of the predicted heat load, and combine the hourly dynamic information of the heat load, the building design heat load, the annual cumulative heat load, the buried pipe parameters and the geothermal boundary parameters into a capacity configuration set;

[0037] According to the heating mode and capacity configuration set, a long-term heat exchange model of a medium-deep underground pipe heat pump system is used to perform dynamic heating simulation to obtain configuration operation data; the long-term heat exchange model of a medium-deep underground pipe heat pump system includes a capacity configuration layer, a heat exchange simulation layer and a parameter optimization layer;

[0038] The system capacity configuration layer determines the system capacity configuration according to the heating mode, building design heat load, annual cumulative heat load and buried pipe parameters; the heat exchange simulation layer conducts long-term heat exchange dynamic simulation of the medium-deep buried pipe according to the heating mode, system capacity configuration, hourly dynamic information of heat load, buried pipe parameters and geothermal boundary parameters to obtain operating temperature data; the parameter optimization layer optimizes the long-term heat exchange model of the medium-deep buried pipe heat pump system according to the configured operating data;

[0039] The project system information to be configured is input into the optimized long-term heat exchange model of the medium-deep buried pipe heat pump system to output the final configuration and operation data of the project system.

[0040] Furthermore, the method for optimizing the long-term heat exchange model of the medium-deep buried pipe heat pump system according to the configuration operation data includes:

[0041] The long-term heat exchange dynamic simulation index is determined by the configuration operation data, and the target optimization function of the long-term heat exchange model of the medium-deep buried pipe heat pump system is determined by the long-term heat exchange dynamic simulation index:

[0042]

[0043] Among them Aim opt is the target optimization function of the long-term heat exchange model of the medium-deep buried pipe heat pump system, c1 is the design heat load weight, c2 is the operating heat load weight, c3 is the operating temperature weight, is the maximum design heat load of the project, is the optimal design heat load of the heat pump system, b is the number of heat load updates, is the building heat load at time i, is the heat load of the heat pump system at time i, is the building dynamic heat load variance, is the temperature decay rate at the water inlet, is the outlet temperature decay rate, is the temperature decay rate of the rock and soil mass, ΔT in is the water inlet temperature decay, ΔT out is the outlet temperature decay, ΔT geo is the temperature attenuation of the rock and soil mass;

[0044] The Monte Carlo method is used to perform random sampling of parameters within the parameter setting range of the long-term heat exchange model of the medium-deep buried pipe heat pump system to generate the model parameter combination p i , i∈[0,r], r is the number of parameter groups, and the loss function value L(p i ) obtains a parameter set and a loss value pair D; the loss function L(p i ) conforms to the Gaussian distribution L(p i )~N(0,U), where U is the element k(pi , p j ), k(p i , p j ) is the kernel function representing the covariance of parameter p;

[0045] Take the current parameter combination as the optimal parameter combination p * , the screening loss function value is greater than L(p * ) parameter combination p i , use the EI function to calculate the value of the selected parameter combination, and select the point with the largest EI function as the most promising set of hyperparameters p new Perform the next evaluation, update the Gaussian distribution, and calculate the Gaussian distribution of the new loss function;

[0046] Repeated sampling selection and update L(p i ) Gaussian distribution operation is performed until the objective optimization function of the long-term heat exchange model of the medium-deep buried pipe heat pump system is minimized, and the final parameter combination is output.

[0047] The beneficial effects of the present invention are:

[0048] The present invention is a capacity configuration design method for a medium-deep buried pipe heat pump system. Compared with the prior art, the present invention has the following technical effects:

[0049] The present invention can improve the accuracy of capacity configuration design through data inversion, model construction, dynamic heating simulation, optimization model, and construction of heating state factors and heating structure factors, thereby improving the speed and accuracy of capacity configuration design of medium-deep buried pipe heat pump systems, greatly saving resources, and improving optimization and adjustment efficiency. It can realize the economy and efficiency of capacity configuration design of medium-deep buried pipe heat pump systems, dynamically optimize and adjust the capacity configuration of medium-deep buried pipe heat pump systems, and provide a guarantee for the accuracy of capacity configuration design of medium-deep buried pipe heat pump systems, which is of great significance to improving system energy efficiency, reducing operating costs, and enhancing user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A flow chart of the steps of a method for designing capacity configuration of a medium-deep buried pipe heat pump system according to the present invention;

[0051] Figure 2A This is the back-calculated fitting data diagram for the first set of data;

[0052] Figure 2B This is the back-calculated fitting data diagram for the second set of data;

[0053] Figure 3 Predicted heat load data graphs for buildings;

[0054] Figure 4AThis is the hourly inlet and outlet water temperature curve of the medium-deep buried pipe;

[0055] Figure 4B It is the underground rock and soil temperature curve of medium and deep buried pipes. DETAILED DESCRIPTION

[0056] The present invention is further described below by means of specific embodiments. The illustrative embodiments and descriptions of the present invention are used to explain the present invention but are not intended to limit the present invention.

[0057] A method for designing capacity configuration of a medium-deep buried pipe heat pump system of the present invention comprises the following steps:

[0058] like Figure 1 As shown, in this embodiment, the following steps are included:

[0059] Obtain short-term heat exchange test data and heat exchange theoretical data of medium-deep buried pipes, and perform data inversion based on the heat exchange test data and the heat exchange theoretical data to obtain comprehensive thermal physical property parameters of rock and soil;

[0060] Obtain historical heat load and historical status data of similar buildings, construct a building heat load prediction model based on the historical heat load, the historical status data and building information, and input the building status information to be predicted into the building heat load prediction model to obtain a predicted heat load;

[0061] Determine the heating mode according to the hourly dynamic information of the predicted heat load, and use the long-term heat exchange model of the medium-deep buried pipe heat pump system to perform dynamic heating simulation to obtain configuration operation data; the configuration operation data includes system capacity configuration and operation temperature data;

[0062] The long-term heat exchange model of the medium-deep buried pipe heat pump system is optimized according to the configuration operation data, and the project system information to be configured is input into the optimized long-term heat exchange model of the medium-deep buried pipe heat pump system to output the project system capacity configuration.

[0063] In this embodiment, the method for obtaining the comprehensive thermophysical property parameters of the rock and soil mass includes:

[0064] Construct medium-deep buried pipe test holes at the project site, and conduct short-term heat exchange tests based on the medium-deep buried pipe test holes to obtain heat exchange test data and original temperature distribution data of the project site;

[0065] According to the original temperature distribution of the project site and the parameters of the buried pipe test hole, a short-term heat exchange simulation of the medium-deep buried pipe was carried out to obtain heat exchange simulation data;

[0066] The heat exchange test data and heat exchange simulation data are back-calculated and fitted, and the comprehensive thermal physical property parameters of the rock and soil are adjusted according to the back-calculation and fitting results. The adjusted comprehensive thermal physical property parameters of the rock and soil are input into the short-term heat exchange simulation model of the medium-deep buried pipe to obtain the adjusted heat exchange simulation data;

[0067] Repeat the above back-calculation fitting and parameter adjustment operations until the back-calculation objective function is minimized, and output the corresponding comprehensive thermal physical property parameters of the rock and soil mass;

[0068] The inverse calculation objective function expression is:

[0069]

[0070] Among them Aim rev is the back-calculation objective function, n is the number of data sets for the heat transfer test, w1=w2=0.2, w3=0.6 are the temperature deviation weights, is the average liquid temperature of the heat transfer test at time i, is the average liquid temperature of the heat exchange simulation at time i, is the water inlet temperature of the heat exchange test at time i, is the water inlet temperature of the heat exchange simulation at time i, is the outlet temperature of the heat exchange test at time i, is the outlet temperature of the heat exchange simulation at time i;

[0071] In the actual evaluation, taking the capacity configuration of the medium-deep buried pipe heat pump system of a large transportation hub project in Changsha as an example, the parameters of the buried pipes constructed on site are shown in Table 1:

[0072] Table 1 Parameters of underground pipes constructed on site

[0073] Physical parameter items Numeric unit Buried pipe type Coaxial sleeve - Casing depth 2500 m Numerical calculation domain depth 3000 m Outer pipe material Oil casing - Outer tube diameter 0.1778 m Outer tube inner diameter 0.1594 m Inner pipe material PE pipe - Inner tube outer diameter 0.11 m Inner tube diameter 0.09 m Circulating fluid water - Thermal conductivity of water 0.618 W / (m·K) Specific heat capacity of water 4.19 kJ / (kg·K) Density of water 995.7 <![CDATA[kg / m 3 ]]> Prandtl number of water 5.42 Dimensionless Initial temperature of the upper surface of the formation 17.027 ℃ Initial geothermal gradient 0.02473 K / m Design flow rate of a single borehole 28 <![CDATA[m 3 / h]]>

[0074] Based on the short-term heat exchange test of the deep underground pipe test hole, the original temperature distribution of the project site was obtained: 19.5℃ at a depth of 100m, 81.6℃ at a depth of 2611m, and the underground rock and soil temperature gradient was 2.47℃ / 100m. The back calculation fitting of two sets of heat exchange test data and heat exchange simulation data was taken as an example. Figure 2A is the back-calculation fitting of the first set of data (the heat exchange test time is from 22:00 on January 12, 2023 to 17:00 on January 23, 2023, and the test time lasts for 260 hours in total). Figure 2B This is the back-calculation fitting of the second set of data (the heat exchange test time is from 19:00 on February 8, 2023 to 21:00 on February 13, 2023, and the test time lasts for 120 hours). The back-calculation fitting results show that the comprehensive specific heat capacity of the rock and soil mass is about 2850 kJ / (m 3·K), the comprehensive thermal conductivity of soil is 2.2W / (m·K)~3.5W / (m·K)2.2W / (m·K)~3.5W / (m·K).

[0075] In this embodiment, the method for obtaining the predicted heat load includes:

[0076] Obtain historical heat load and historical status data of similar buildings, input the historical status data into the state function to obtain the heating state factor, and construct the heating structure factor according to the building information;

[0077] The historical heat load, historical state data, heating state factor and heating structure factor are combined to obtain a comprehensive prediction set, which is divided into a training set and a test set according to a ratio of 7:3;

[0078] Construct a building heat load prediction model, which includes genetic algorithm, BP neural network, loss function and grid search algorithm;

[0079] The genetic algorithm optimizes the initial weights and biases of the BP neural network; the BP neural network learns the highly nonlinear relationship between historical heat load, historical state data, heating state factor and heating structure factor, and predicts the building heat load based on the input state data, heating state factor and heating structure factor; the grid search algorithm tunes the hyperparameters of the building heat load prediction model; the loss function measures the difference between the predicted value and the true value, and the expression is:

[0080]

[0081] Among them, α = 0.5, β = 0.5 are the weights of the loss function, δ is the error hyperparameter used to control the error threshold, Q i is the true value of heat load in stage i, is the predicted value of heat load in stage i, and m is the number of samples;

[0082] Input the building status information to be predicted (the planned building area is about 1.15 million square meters, and the building heating period is from December 1 to February 28 of the following year) into the building heat load prediction model to obtain the predicted heat load as follows Figure 3 As shown, the project's designed thermal load is 37,992 kW and the annual cumulative thermal load is 18,675,800 kWh / a.

[0083] In this embodiment, the method for obtaining the comprehensive prediction set includes:

[0084] Input the historical state data into the state function to obtain the heating state factor, the expression is:

[0085]

[0086] Among them Ista，i is the heating state factor of stage i, u1 is the temperature weight, u2 is the level weight, T ei is the outdoor ambient temperature in stage i, T hi is the indoor temperature of the historical building in stage i, T ni is the indoor temperature of the proposed building in phase i, H0 is the standard humidity, H hi is the indoor humidity of the historical building in stage i, H ni is the indoor humidity of the proposed building in phase i, is the heating grade of the proposed building, Heating rating for historic buildings;

[0087] The heating structure factor is constructed based on the building information, and the expression is:

[0088]

[0089] Among them I str is the heating structure factor, u3 is the space weight, u4 is the insulation weight, A n is the planned building area, V n is the volume of the proposed building space, High n A is the proposed building height, h is the historical building area, V h For the volume of historical building space, High h For the historical building, is the insulation grade of the proposed building, R n is the air renewal rate of the proposed building, is the insulation level of historical buildings, R h Air renewal rate for historic buildings;

[0090] The historical heat load, historical status data, heating status factors and heating structure factors are combined to obtain a comprehensive prediction set.

[0091] In this embodiment, the method for determining the heating mode includes:

[0092] Obtain hourly dynamic information of predicted heat load, define heat load update as a heat load change greater than 5% of design heat load, calculate heat load update dynamic index, set heat load update dynamic index threshold, compare heat load update dynamic index with corresponding heat load update dynamic index threshold to determine heating mode; the heat load dynamic index includes update number, average update frequency, heat load update standard deviation, heat load update deviation, and heat load update variation coefficient;

[0093] The method for determining the heating mode is specifically as follows: when more than 1 / 2 of the dynamic heat load index exceeds the corresponding heat load dynamic index threshold, dynamic heating is performed, otherwise constant heating is performed; the specific scheme of the dynamic heating is that the medium-deep buried pipe heat pump system bears all the dynamic heat loads of the building; the specific scheme of the constant heating is that the medium-deep buried pipe heat pump system bears the constant heat load of the building foundation, and the remaining heat load is borne by the auxiliary heat source; the constant heat load is determined according to the optimal heating efficiency of the medium-deep buried pipe;

[0094] In the actual evaluation, the dynamic indicators of heat load update of a planned airport project in Changsha are: update number 10 times, average update frequency 36.5 days / time, heat load update standard deviation 482.2kW, heat load update deviation absolute value 2100kW, heat load update coefficient of variation 0.228. The corresponding heat load dynamic indicator thresholds are: update number greater than 15 times, average update frequency less than 30 days / time, heat load update standard deviation greater than 600kW, heat load update deviation absolute value greater than 2000kW, heat load update coefficient of variation greater than 0.2. Among them, only the absolute value of heat load update deviation and heat load update coefficient of variation are greater than the corresponding indicator thresholds, so the heating mode is constant heating.

[0095] In this embodiment, the method for obtaining the configuration operation data includes:

[0096] Determine the building design heat load and the annual cumulative heat load based on the hourly dynamic information of the predicted heat load;

[0097] The hourly dynamic information of heat load, building design heat load, annual cumulative heat load, buried pipe parameters and geothermal boundary parameters are combined into a capacity configuration set; the geothermal boundary parameters include the original temperature distribution data of the project site and the comprehensive thermal physical property parameters of the rock and soil mass;

[0098] According to the heating mode and capacity configuration set, a long-term heat exchange model of a medium-deep underground pipe heat pump system is used to perform dynamic heating simulation to obtain configuration operation data; the long-term heat exchange model of a medium-deep underground pipe heat pump system includes a capacity configuration layer, a heat exchange simulation layer and a parameter optimization layer;

[0099] The system capacity configuration layer determines the system capacity configuration according to the heating mode, the building design heat load, the annual cumulative heat load and the buried pipe parameters; the system capacity configuration includes the number of buried pipe holes, the heat load of a single buried pipe and the annual cumulative heat load;

[0100] The heat exchange simulation layer conducts long-term heat exchange dynamic simulation of deep-layer underground pipes according to the heating mode, system capacity configuration, hourly dynamic information of heat load, underground pipe parameters and geothermal boundary parameters to obtain operating temperature data;

[0101] The parameter optimization layer optimizes the long-term heat exchange model of the medium-deep buried pipe heat pump system according to the configuration operation data;

[0102] The project system information to be configured is input into the optimized long-term heat exchange model of the medium-deep buried pipe heat pump system to output the final configuration and operation data of the project system.

[0103] In this embodiment, the method for optimizing the long-term heat exchange model of the medium-deep buried pipe heat pump system according to the configuration operation data includes:

[0104] The long-term heat exchange dynamic simulation index is determined by the configuration operation data, and the target optimization function of the long-term heat exchange model of the medium-deep buried pipe heat pump system is determined by the long-term heat exchange dynamic simulation index:

[0105]

[0106] Among them Aim opt is the target optimization function of the long-term heat exchange model of the medium-deep buried pipe heat pump system, c1 is the design heat load weight, c2 is the operating heat load weight, c3 is the operating temperature weight, is the maximum design heat load of the project, is the optimal design heat load of the heat pump system, b is the number of heat load updates, is the building heat load at time i, is the heat load of the heat pump system at time i, is the building dynamic heat load variance, is the temperature decay rate at the water inlet, is the outlet temperature decay rate, is the temperature decay rate of the rock and soil mass, ΔT in is the water inlet temperature decay, ΔT out is the outlet temperature decay, ΔT geo is the temperature attenuation of the rock and soil mass;

[0107] The Monte Carlo method is used to perform random sampling of parameters within the parameter setting range of the long-term heat exchange model of the medium-deep buried pipe heat pump system to generate the model parameter combination p i , i∈[0,r], r is the number of parameter groups, and the loss function value L(p i ) obtains a parameter set and a loss value pair D; the loss function L(p i ) conforms to the Gaussian distribution L(p i )~N(0,U), where U is the element k(p i , p j ), k(p i , p j ) is the kernel function representing the covariance of parameter p;

[0108] Take the current parameter combination as the optimal parameter combination p * , the screening loss function value is greater than L(p * ) parameter combination p i , use the EI function to calculate the value of the selected parameter combination, and select the point with the largest EI function as the most promising set of hyperparameters p new Conduct the next assessment;

[0109] Update the Gaussian distribution and calculate the Gaussian distribution of the new loss function:

[0110]

[0111] k′=[k(p new ,p1),…,k(p newi , p r )]

[0112] Where L(p new ) is the loss function value of the most promising set of hyperparameters, k′ is the most promising parameter combination p new Combined with parameter p i The covariance set, P[L(p new |D,p new )] is the new loss function distribution, μ p = k′ T U -1 L(p i ) is the mean of the updated loss function value, To update the variance of the loss function value;

[0113] Repeated sampling selection and update L(p i ) Gaussian distribution operation is performed until the target optimization function of the long-term heat exchange model of the medium-deep buried pipe heat pump system is minimized, and the final parameter combination is output;

[0114] In the actual evaluation, the heating mode of a certain airport project in Changsha was selected as constant heating. The corresponding predicted heat load, buried pipe parameters and geothermal boundary parameters were input into the optimized long-term heat exchange model of the medium-deep buried pipe heat pump system to output the final configuration and operation data of the project system. The configuration design of the medium-deep buried pipe heat pump system is shown in Table 2. The hourly inlet and outlet water temperature curves of the medium-deep buried pipe are shown in Table 2. Figure 4A As shown in the figure, the underground rock and soil temperature curve of the medium-deep buried pipe is as follows Figure 4B As shown;

[0115] Table 2 Configuration table of medium and deep buried pipe heat pump system

[0116] Parameters Numeric unit Heating season time December 1st to February 28th - Heating season duration 90 d A single medium-deep pipe can bear the continuous and constant heat load of the building. 360 kW The cumulative load of a building borne by a single medium-deep pipe during the heating season 77.76 10,000 kWh / a Number of buried pipe holes in the middle and deep layers under cumulative heat load 24 indivual Minimum inlet water temperature on the ground source side 5.74 ℃ Minimum outlet water temperature at the ground source side 14.95 ℃

[0117] Under the current calculation conditions, the deep buried pipe heat pump system in a certain airport project in Changsha only needs to be equipped with 24 buried pipes. A single buried pipe bears a constant heat load of 360kW, and can bear a cumulative heat load of 777,600 kWh / a throughout the year. The minimum inlet water temperature of the medium-deep buried pipe is 5.74℃, which is greater than the minimum threshold of 5℃; the minimum outlet water temperature of the buried pipe is 14.95℃; therefore, under the current calculation scenario, the inlet and outlet water temperatures on the ground source side of the medium-deep buried pipe are generally good, the heat pump unit can operate stably and efficiently all year round, and the medium-deep buried pipe ground source heat pump system has stable and reliable heating and good heating endurance ; If the medium-deep buried pipe ground source heat pump system is used for dynamic heating, it is calculated by the long-term heat exchange model of the medium-deep buried pipe heat pump system that 65 buried pipes are needed, and the construction cost is higher. The design heat load of a single buried pipe is 584.49kW, and the cumulative annual heat load is 287,300 kWh / a. The design heat load is much greater than the constant heating situation and the cumulative annual heat load is less than the constant heating situation, and the heating efficiency is lower. The selection of heating mode and the capacity configuration design of the buried pipe ground source heat pump system can reduce the construction cost of the medium-deep buried pipe ground source heat pump system and improve the heating efficiency.

[0118] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for designing capacity configuration of a medium-deep buried pipe heat pump system, characterized in that: The following steps are involved: S1. Obtain short-term heat exchange test data and heat exchange theoretical data of medium-deep buried pipes, and perform data inversion based on the heat exchange test data and the heat exchange theoretical data to obtain comprehensive thermal physical property parameters of rock and soil; S2. Obtain historical heat load and historical status data of similar buildings, construct a building heat load prediction model based on the historical heat load, the historical status data and building information, and input the building status information to be predicted into the building heat load prediction model to obtain a predicted heat load; S3, determining the heating mode according to the hourly dynamic information of the predicted heat load, and using the long-term heat exchange model of the medium-deep buried pipe heat pump system to perform dynamic heating simulation to obtain configuration operation data; the configuration operation data includes system capacity configuration and operation temperature data; S4. Optimize the long-term heat exchange model of the medium-deep buried pipe heat pump system according to the configuration operation data, input the project system information to be configured into the optimized long-term heat exchange model of the medium-deep buried pipe heat pump system to output the project system capacity configuration.

2. According to claim 1, a method for designing capacity configuration of a medium-deep buried pipe heat pump system is characterized in that: The method for obtaining the comprehensive thermal physical property parameters of the rock and soil mass comprises: Construct medium-deep buried pipe test holes at the project site, and conduct short-term heat exchange tests based on the medium-deep buried pipe test holes to obtain heat exchange test data and original temperature distribution data of the project site; According to the original temperature distribution of the project site and the parameters of the buried pipe test hole, a short-term heat exchange simulation of the medium-deep buried pipe was carried out to obtain heat exchange simulation data; The heat exchange test data and heat exchange simulation data are back-calculated and fitted, and the comprehensive thermal physical property parameters of the rock and soil are adjusted according to the back-calculation and fitting results. The adjusted comprehensive thermal physical property parameters of the rock and soil are input into the short-term heat exchange simulation model of the medium-deep buried pipe to obtain the adjusted heat exchange simulation data; Repeat the above back-calculation fitting and parameter adjustment operations until the back-calculation objective function is minimized, and output the corresponding comprehensive thermal physical properties parameters of the rock and soil mass.

3. According to claim 1, a method for designing capacity configuration of a medium-deep buried pipe heat pump system is characterized in that: The method for obtaining the predicted heat load comprises: Obtain historical heat load and historical status data of similar buildings, input the historical status data into the state function to obtain the heating state factor, and construct the heating structure factor according to the building information; The historical heat load, historical status data, heating status factor and heating structure factor are combined to obtain a comprehensive prediction set, and the comprehensive prediction set is divided into a training set and a test set according to a ratio of 7:3; Construct a building heat load prediction model, which includes genetic algorithm, BP neural network, loss function and grid search algorithm; The genetic algorithm optimizes the initial weights and biases of the BP neural network; the BP neural network learns the highly nonlinear relationship between historical heat load, historical state data, heating state factor and heating structure factor, and predicts the building heat load based on the input state data, heating state factor and heating structure factor; the grid search algorithm tunes the hyperparameters of the building heat load prediction model; the loss function measures the difference between the predicted value and the true value, and the expression is: Among them, α and β are the weights of the loss function, δ is the error hyperparameter used to control the error threshold, and Q i is the true value of heat load in stage i, is the predicted value of heat load in stage i, and m is the number of samples The state information of the building to be predicted is input into the building heat load prediction model to obtain the predicted heat load.

4. According to claim 3, a method for designing capacity configuration of a medium-deep buried pipe heat pump system is characterized in that: The method for obtaining the comprehensive prediction set comprises: Input the historical state data into the state function to obtain the heating state factor, the expression is: Among them I sta,i is the heating state factor of stage i, u1 is the temperature weight, u2 is the level weight, T ei is the outdoor ambient temperature in stage i, T hi is the indoor temperature of the historical building in stage i, T ni is the indoor temperature of the proposed building in phase i, H0 is the standard humidity, H hi is the indoor humidity of the historical building in stage i, H ni is the indoor humidity of the proposed building in phase i, is the heating grade of the proposed building, Heating rating for historic buildings; The heating structure factor is constructed based on the building information, and the expression is: Among them I str is the heating structure factor, u3 is the space weight, u4 is the insulation weight, A n is the planned building area, V n is the volume of the proposed building space, High n A is the proposed building height, h is the historical building area, V h For the volume of historical building space, High h For the historical building, is the insulation grade of the proposed building, R n is the air renewal rate of the proposed building, is the insulation level of historical buildings, R h Air renewal rate for historic buildings; The historical heat load, historical status data, heating status factors and heating structure factors are combined to obtain a comprehensive prediction set.

5. According to claim 1, a method for designing capacity configuration of a medium-deep buried pipe heat pump system is characterized in that: The method for determining the heating mode comprises: Obtain hourly dynamic information of predicted heat load, define heat load change greater than 5% of design heat load as a heat load update, calculate heat load update dynamic index, set heat load update dynamic index threshold, compare heat load update dynamic index with corresponding heat load update dynamic index threshold to determine heating mode; The method for determining the heating mode is specifically as follows: dynamic heating is performed when more than 1 / 2 of the heat load dynamic indicators exceed the corresponding heat load dynamic indicator threshold, otherwise constant heating is performed; the specific plan for dynamic heating is that the medium-deep buried pipe heat pump system bears all the dynamic heat loads of the building; the specific plan for constant heating is that the medium-deep buried pipe heat pump system bears the constant heat load of the building foundation, and the remaining heat load is borne by the auxiliary heat source; the constant heat load is determined according to the optimal heating efficiency of the medium-deep buried pipes.

6. According to claim 1, a method for designing capacity configuration of a medium-deep buried pipe heat pump system is characterized in that: The method for obtaining the configuration operation data comprises: Determine the building design heat load and the annual cumulative heat load based on the hourly dynamic information of the predicted heat load, and combine the hourly dynamic information of the heat load, the building design heat load, the annual cumulative heat load, the buried pipe parameters and the geothermal boundary parameters into a capacity configuration set; According to the heating mode and capacity configuration set, a long-term heat exchange model of a medium-deep underground pipe heat pump system is used to perform dynamic heating simulation to obtain configuration operation data; the long-term heat exchange model of a medium-deep underground pipe heat pump system includes a capacity configuration layer, a heat exchange simulation layer and a parameter optimization layer; The system capacity configuration layer determines the system capacity configuration according to the heating mode, building design heat load, annual cumulative heat load and buried pipe parameters; the heat exchange simulation layer conducts long-term heat exchange dynamic simulation of the medium-deep buried pipe according to the heating mode, system capacity configuration, hourly dynamic information of heat load, buried pipe parameters and geothermal boundary parameters to obtain operating temperature data; the parameter optimization layer optimizes the long-term heat exchange model of the medium-deep buried pipe heat pump system according to the configured operating data; The project system information to be configured is input into the optimized long-term heat exchange model of the medium-deep buried pipe heat pump system to output the final configuration and operation data of the project system.

7. According to claim 1, a method for designing capacity configuration of a medium-deep buried pipe heat pump system is characterized in that: The method for optimizing the long-term heat exchange model of the medium-deep buried pipe heat pump system according to the configuration operation data comprises: The long-term heat exchange dynamic simulation index is determined by the configuration operation data, and the target optimization function of the long-term heat exchange model of the medium-deep buried pipe heat pump system is determined by the long-term heat exchange dynamic simulation index: Among them Aim opt is the target optimization function of the long-term heat exchange model of the medium-deep buried pipe heat pump system, c1 is the design heat load weight, c2 is the operating heat load weight, c3 is the operating temperature weight, is the maximum design heat load of the project, is the optimal design heat load of the heat pump system, b is the number of heat load updates, is the building heat load at time i, is the heat load of the heat pump system at time i, V is the building dynamic heat load variance, Tin is the temperature decay rate at the water inlet, is the outlet temperature decay rate, is the temperature decay rate of the rock and soil mass, ΔT in is the water inlet temperature decay, ΔT out is the outlet temperature decay, ΔT geo is the temperature attenuation of the rock and soil mass; The Monte Carlo method is used to perform random sampling of parameters within the parameter setting range of the long-term heat exchange model of the medium-deep buried pipe heat pump system to generate the model parameter combination p i , i∈[0,r], r is the number of parameter groups, and the loss function value L(p i ) obtains a parameter set and a loss value pair D; the loss function L(p i ) conforms to the Gaussian distribution L(p i )~N(0,U), U is the element k(p i ,p j ), k(p i ,p j ) is the kernel function representing the covariance of parameter p; Take the current parameter combination as the optimal parameter combination p * , the screening loss function value is greater than L(p * ) parameter combination p i , use the EI function to calculate the value of the selected parameter combination, and select the point with the largest EI function as the most promising set of hyperparameters p new Perform the next evaluation, update the Gaussian distribution, and calculate the Gaussian distribution of the new loss function; Repeated sampling selection and update L(p i ) Gaussian distribution operation is performed until the objective optimization function of the long-term heat exchange model of the medium-deep buried pipe heat pump system is minimized, and the final parameter combination is output.

Images