Heat source tower heat pump applicability evaluation model construction method and system and applicability evaluation method and system

Abstract

The invention provides a heat source tower heat pump applicability evaluation model construction and applicability evaluation method and system. The heat source tower heat pump applicability evaluation model construction method comprises the steps that historical state data of a heat source tower heat pump, all index values of the historical state data under a preset applicability evaluation index system and real applicability evaluation results corresponding to all the index values are obtained; taking each index value of the historical state data under the applicability evaluation index system as input data of training data; taking a real applicability evaluation result corresponding to each index value as output data of the training data; based on the input data and the output data, training a neural network model by using a particle swarm algorithm to obtain an applicability evaluation model; the constructed applicability evaluation model is based on the applicability evaluation index system, the subjectivity of manual judgment can be reduced, the neural network model is trained by adopting the particle swarm optimization, the fitting precision of the applicability evaluation model can be improved, and the accuracy of model prediction can be improved.

Description

Technical Field

[0001] The present invention relates to the field of thermal energy technology, and in particular to the construction of a heat source tower heat pump applicability evaluation model, a applicability evaluation method and a system. Background Art

[0002] At present, with the continuous increase in building area, building energy consumption has increased dramatically. The contradiction between the growing building energy consumption and limited energy needs to be resolved urgently, and the importance of energy efficiency and renewable energy has become increasingly prominent. Heat source tower heat pump technology is an advanced way of energy utilization. It can extract heat energy from a low-temperature environment and store it through a heat source tower, and release energy during peak energy consumption periods, thereby achieving efficient utilization and storage of energy. Heat source tower heat pump technology is an important way to achieve safe, economical, green and efficient operation of energy systems. It is urgent to promote the further development of heat source tower heat pump technology to reduce energy consumption and reduce dependence on fossil fuels, and promote sustainable development in the energy field.

[0003] Heat source tower heat pump technology provides a new idea for low-carbon and efficient operation of energy systems. However, the current heat source tower heat pump technology is still not mature enough, the construction and maintenance costs are high, and the climate and regional differences have a great impact on the energy conversion efficiency of heat source tower heat pumps. At present, most of the research on heat source tower heat pump systems focuses on the optimization and efficiency improvement of the system. The existing heat source tower heat pump applicability evaluation method is affected by subjective factors, which makes the constructed applicability evaluation model less universal, resulting in low accuracy of the applicability evaluation results. Summary of the invention

[0004] In order to solve the problem that the applicability evaluation model constructed in the prior art for heat source tower heat pump is weak in universality due to the influence of subjective factors, resulting in low accuracy of the applicability evaluation result, the present invention proposes a method for constructing a heat source tower heat pump applicability evaluation model, including:

[0005] Obtaining historical status data of a heat source tower heat pump, various index values ​​of the historical status data under a preset applicability evaluation index system, and actual applicability evaluation results corresponding to the various index values;

[0006] Using each index value of the historical state data under the applicability evaluation index system as input data of training data;

[0007] Using the true applicability evaluation results corresponding to the index values ​​as output data of training data;

[0008] Based on the input data and the output data, the neural network model is trained using a particle swarm algorithm to obtain a suitability evaluation model.

[0009] Optionally, the using a particle swarm algorithm to train a neural network model based on the input data and the output data to obtain a suitability evaluation model includes:

[0010] Based on the input data, output data, randomly generated initial weights and initial thresholds, a neural network model is trained using a particle swarm algorithm to obtain optimal weights and optimal thresholds;

[0011] According to the optimal weight and the optimal threshold, a suitability evaluation model is obtained.

[0012] Optionally, the using of a particle swarm algorithm to train the neural network model to obtain an optimal weight and an optimal threshold comprises:

[0013] Initialize the parameters of the particle swarm algorithm, including the number of particles in the particle swarm, the particle position and the number of iterations, as well as initialize the individual extreme value and the global optimal extreme value, and the initial position of each particle corresponds to a set of initial weights and initial thresholds of the neural network model;

[0014] Inputting the input data of the training data into the neural network model to obtain prediction data corresponding to the input data;

[0015] Calculating the fitness function value of each particle based on the square of the difference between the predicted data and the output data;

[0016] Based on the fitness function value of each particle, the particle swarm algorithm is used to iteratively update the individual extreme value and the global optimal extreme value until the updated individual extreme value and the global optimal extreme value reach the convergence condition. At this time, the preset inertia weight update expression is used to update the initial weight and the initial threshold, and the updated initial weight is determined as the optimal weight, and the updated initial threshold is determined as the optimal threshold.

[0017] Optionally, the adopting a preset inertia weight update expression to update the initial weight and the initial threshold includes:

[0018] The inertia weight of the particle is updated using a preset inertia weight update expression to obtain an updated inertia weight;

[0019] It is determined whether the maximum number of iterations has been reached. If so, the initial weight and the initial threshold are updated based on the updated inertia weight.

[0020] Optionally, updating the initial weight and the initial threshold based on the updated inertia weight includes:

[0021] Obtaining a learning rate of the neural network model according to the updated inertia weight;

[0022] According to the learning rate of the neural network model, the initial weight and initial threshold are updated through a back propagation algorithm.

[0023] Optionally, the inertia weight update expression is as follows:

[0024]

[0025] Among them, ω(t) represents the inertia weight at the tth iteration; ω max Represents the maximum value of inertia weight; ω min represents the minimum value of the inertia weight; t represents the current number of iterations; t max Represents the maximum number of iterations; e is a natural constant.

[0026] Optionally, the update expression of the initial weight is as follows:

[0027] w(t+1)=w(t)+η(P best (t)-w(t))+η(G best (t)-w(t));

[0028] The update expression of the initial threshold is as follows:

[0029]

[0030] Wherein, w(t+1) represents the initial weight of the neural network model at the t+1th iteration; w(t) represents the initial weight of the neural network model at the tth iteration; η represents the learning rate of the neural network model; P best (t) represents the individual extreme value at the tth iteration; G best (t) represents the global optimal extreme value at the t-th iteration; b(t+1) represents the initial threshold at the t+1-th iteration; b(t) represents the initial threshold at the t-th iteration; represents the output data at the t-th iteration; Represents the predicted data corresponding to the input data at the tth iteration.

[0031] Optionally, the neural network model is a back-propagation neural network model.

[0032] Based on the same inventive concept, the present invention also provides a heat source tower heat pump applicability evaluation model construction system, comprising:

[0033] A data acquisition module, used to acquire historical status data of a heat source tower heat pump, various index values ​​of the historical status data under a preset applicability evaluation index system, and actual applicability evaluation results corresponding to the various index values;

[0034] A training input module, used to use the index values ​​of the historical status data under the applicability evaluation index system as input data of training data;

[0035] A training output module, used to use the true applicability evaluation results corresponding to the indicator values ​​as output data of training data;

[0036] The model building module is used to train the neural network model based on the input data and the output data using a particle swarm algorithm to obtain a suitability evaluation model.

[0037] Optionally, the model building module includes:

[0038] A model training submodule is used to train the neural network model using a particle swarm algorithm based on the input data, output data, randomly generated initial weights and initial thresholds to obtain optimal weights and optimal thresholds;

[0039] The parameter adjustment submodule is used to obtain a suitability evaluation model according to the optimal weight and the optimal threshold.

[0040] Optionally, the model training submodule includes:

[0041] A parameter initialization unit, used to initialize the parameters of the particle swarm algorithm, the parameters including the number of particles in the particle swarm, the particle positions and the number of iterations, as well as initializing individual extreme values ​​and global optimal extreme values, and the initial position of each particle corresponds to a set of initial weights and initial thresholds of the neural network model;

[0042] A model output unit, used to input the input data of the training data into the neural network model to obtain prediction data corresponding to the input data;

[0043] A fitness calculation unit, used to calculate the fitness function value of each particle based on the square of the difference between the predicted data and the output data;

[0044] A parameter updating unit is used to iteratively update the individual extreme value and the global optimal extreme value based on the fitness function value of each particle by using the particle swarm algorithm until the updated individual extreme value and the global optimal extreme value reach the convergence condition. At this time, a preset inertia weight update expression is used to update the initial weight and the initial threshold, and the updated initial weight is determined as the optimal weight, and the updated initial threshold is determined as the optimal threshold.

[0045] Optionally, the parameter updating unit includes:

[0046] The weight updating subunit is used to update the inertia weight of the particle using a preset inertia weight updating expression to obtain an updated inertia weight;

[0047] The parameter determination subunit is used to determine whether the maximum number of iterations has been reached. If so, the initial weight and the initial threshold are updated based on the updated inertia weight.

[0048] Optionally, the parameter determination subunit includes:

[0049] A learning rate updating component, used to obtain a learning rate of the neural network model according to the updated inertia weight;

[0050] The optimal parameter updating component is used to update the initial weight and initial threshold according to the learning rate of the neural network model through a back propagation algorithm.

[0051] Optionally, the inertia weight update expression is as follows:

[0052]

[0053] Among them, ω(t) represents the inertia weight at the tth iteration; ω max Represents the maximum value of inertia weight; ω min represents the minimum value of the inertia weight; t represents the current number of iterations; t max Represents the maximum number of iterations; e is a natural constant.

[0054] Optionally, the update expression of the initial weight is as follows:

[0055] w(t+1)=w(t)+η(P best (t)-w(t))+η(G best (t)-w(t));

[0056] The update expression of the initial threshold is as follows:

[0057]

[0058] Wherein, w(t+1) represents the initial weight of the neural network model at the t+1th iteration; w(t) represents the initial weight of the neural network model at the tth iteration; η represents the learning rate of the neural network model; P best (t) represents the individual extreme value at the tth iteration; G best (t) represents the global optimal extreme value at the t-th iteration; b(t+1) represents the initial threshold at the t+1-th iteration; b(t) represents the initial threshold at the t-th iteration; represents the output data at the t-th iteration; Represents the predicted data corresponding to the input data at the tth iteration.

[0059] Optionally, the neural network model is a back-propagation neural network model.

[0060] Based on the same inventive concept, the present invention also provides a method for evaluating the suitability of a heat source tower heat pump, comprising:

[0061] Get the status data of the heat pump in the heat source tower;

[0062] According to the state data, using a suitability evaluation model to evaluate the suitability of the heat source tower heat pump, to obtain a suitability evaluation result of the heat source tower heat pump;

[0063] Wherein, the applicability evaluation model is constructed using the aforementioned method for constructing a heat source tower heat pump applicability evaluation model.

[0064] Optionally, the comprehensive evaluation index system of applicability includes: system performance index, energy-saving and environmental protection performance index and economic index;

[0065] The system performance index includes one or more of the following: cooling capacity of the heat source tower heat pump, heating capacity of the heat source tower heat pump, winter heating performance coefficient of the heat source tower heat pump host, summer cooling performance coefficient of the heat source tower heat pump host and system comprehensive performance coefficient;

[0066] The energy-saving and environmental protection index includes one or more of the following: carbon emission cost and system energy consumption coefficient;

[0067] The economic indicators include one or more of the following: investment cost and operating cost;

[0068] The status data includes one or more of the following: heating capacity, cooling capacity, air mass flow, water temperature on the unit outlet side, water temperature on the unit inlet side, host input power, compressor power consumption, tower pump side power consumption, tower side fan power consumption, fossil fuel consumption, basic carbon content of fossil fuels, carbon oxidation rate of fossil fuels, relative molecular mass ratio of carbon dioxide to carbon, power consumption of air-conditioning system, energy consumption, cooling demand of associated buildings, heating demand of associated buildings, equipment procurement cost, installation cost, land cost, construction cost, host cost, chiller cost, solution pump cost, operating solution cost and water cost.

[0069] Based on the same inventive concept, the present invention also provides a heat source tower heat pump applicability evaluation system, comprising:

[0070] The operation data acquisition module is used to obtain the status data of the heat pump of the heat source tower;

[0071] A suitability evaluation module, used to evaluate the suitability of the heat source tower heat pump according to the state data using a suitability evaluation model to obtain a suitability evaluation result of the heat source tower heat pump;

[0072] Wherein, the applicability evaluation model is constructed using the aforementioned method for constructing a heat source tower heat pump applicability evaluation model.

[0073] Optionally, the comprehensive evaluation index system of applicability includes: system performance index, energy-saving and environmental protection performance index and economic index;

[0074] The system performance index includes one or more of the following: cooling capacity of the heat source tower heat pump, heating capacity of the heat source tower heat pump, winter heating performance coefficient of the heat source tower heat pump host, summer cooling performance coefficient of the heat source tower heat pump host and system comprehensive performance coefficient;

[0075] The energy-saving and environmental protection index includes one or more of the following: carbon emission cost and system energy consumption coefficient;

[0076] The economic indicators include one or more of the following: investment cost and operating cost;

[0077] The status data includes one or more of the following: heating capacity, cooling capacity, air mass flow, water temperature on the unit outlet side, water temperature on the unit inlet side, host input power, compressor power consumption, tower pump side power consumption, tower side fan power consumption, fossil fuel consumption, basic carbon content of fossil fuels, carbon oxidation rate of fossil fuels, relative molecular mass ratio of carbon dioxide to carbon, power consumption of air-conditioning system, energy consumption, cooling demand of associated buildings, heating demand of associated buildings, equipment procurement cost, installation cost, land cost, construction cost, host cost, chiller cost, solution pump cost, operating solution cost and water cost.

[0078] In another aspect, the present invention further provides an electronic device, comprising: at least one processor and a memory; the memory and the processor are connected via a bus;

[0079] The memory is used to store one or more programs;

[0080] When the one or more programs are executed by the at least one processor, a method for constructing a heat source tower heat pump suitability evaluation model or a method for evaluating the suitability of a heat source tower heat pump as described above is implemented.

[0081] On the other hand, the present invention also provides a computer-readable storage medium having an execution program stored thereon, which, when executed, implements a method for constructing a heat source tower heat pump suitability evaluation model or a heat source tower heat pump suitability evaluation system as described above.

[0082] Compared with the closest prior art, the present invention has the following beneficial effects:

[0083] 1. The present invention provides a method and system for constructing a suitability evaluation model for a heat source tower heat pump, comprising: obtaining historical status data of a heat source tower heat pump, various index values ​​of the historical status data under a preset suitability evaluation index system, and actual suitability evaluation results corresponding to the various index values; using the various index values ​​of the historical status data under the suitability evaluation index system as input data of training data; using the actual suitability evaluation results corresponding to the various index values ​​as output data of training data; based on the input data and the output data, training a neural network model using a particle swarm algorithm to obtain a suitability evaluation model; when constructing a suitability evaluation model, the present invention can reduce the subjectivity of human judgment based on the suitability evaluation index system, so that the evaluation results are more objective; and by training the neural network model using a particle swarm algorithm, the fitting accuracy of the suitability evaluation model can be improved, and the accuracy of model prediction can be increased; thereby, the suitability evaluation model constructed by the method of the present invention can not only reduce the influence of subjective factors in the suitability evaluation process, but also improve the accuracy of the suitability evaluation results of the heat source tower heat pump.

[0084] 2. The present invention also provides a method and system for evaluating the suitability of a heat source tower heat pump, comprising: obtaining status data of the heat source tower heat pump; based on the status data, using a suitability evaluation model to evaluate the suitability of the heat source tower heat pump, and obtaining a suitability evaluation result of the heat source tower heat pump; the present invention uses a suitability evaluation model constructed based on a suitability evaluation index system, a particle swarm algorithm and a neural network model to evaluate the suitability of a heat source tower heat pump, which can reduce the subjectivity of human judgment, make the evaluation result more objective, and increase the accuracy of the suitability evaluation of the heat source tower heat pump. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 A schematic flow chart of a method for constructing a heat source tower heat pump suitability evaluation model provided by the present invention;

[0086] Figure 2 A schematic diagram of a process for optimizing a neural network model using a particle swarm algorithm in a heat source tower heat pump applicability evaluation model construction method provided by the present invention;

[0087] Figure 3 A schematic diagram of the structural composition of a heat source tower heat pump applicability evaluation model construction system provided by the present invention;

[0088] Figure 4 A schematic flow chart of a heat source tower heat pump suitability evaluation method provided by the present invention;

[0089] Figure 5A schematic diagram of the execution flow of a suitability evaluation model constructed based on a particle swarm algorithm and a neural network in a heat source tower heat pump suitability evaluation method provided by the present invention;

[0090] Figure 6 A schematic diagram of the structural composition of a heat source tower heat pump applicability evaluation system provided by the present invention;

[0091] Figure 7 A schematic diagram of an electronic device provided by the present invention. DETAILED DESCRIPTION

[0092] The present invention provides a heat source tower heat pump suitability evaluation model construction, suitability evaluation method and system. The specific implementation methods of the present invention are further described in detail below in conjunction with the accompanying drawings.

[0093] Embodiment 1:

[0094] The present invention provides a heat source tower heat pump applicability model construction method, the flow diagram is as follows Figure 1 As shown, including:

[0095] Step 1: Obtain historical status data of the heat source tower heat pump, various index values ​​of the historical status data under a preset applicability evaluation index system, and actual applicability evaluation results corresponding to various index values;

[0096] Step 2: Use the index values ​​of historical state data under the applicability evaluation index system as the input data of training data;

[0097] Step 3: Use the actual applicability evaluation results corresponding to each indicator value as the output data of the training data;

[0098] Step 4: Based on the input data and output data, the neural network model is trained using the particle swarm optimization (PSO) to obtain a suitability evaluation model.

[0099] Generally, the energy conversion efficiency of a heat source tower heat pump is related to many factors, such as heat source temperature, compressor efficiency, external climate and environmental conditions, regional differences, and the design and type of the heat pump, which results in the existing heat source tower heat pump applicability evaluation model being less universal and the evaluation indicators being less comprehensive when evaluating the applicability of the heat source tower heat pump. Therefore, when constructing the applicability evaluation model of the heat source tower heat pump, the historical status data used in the present invention takes into account many factors, such as: heating capacity, cooling capacity, air mass flow, water temperature at the water outlet side of the unit, water temperature at the water inlet side of the unit, etc. Temperature, host input power, compressor power consumption, tower pump side power consumption, tower side fan power consumption, fossil fuel consumption, basic carbon content of fossil fuels, carbon oxidation rate of fossil fuels, relative molecular mass ratio of carbon dioxide to carbon, power consumption of air conditioning system, energy consumption, cooling demand of related buildings, heating demand of related buildings, equipment procurement cost, installation cost, land cost, construction cost, host cost, chiller cost, solution pump cost, operating solution cost and water cost, and considering the impact of climate and regional differences on the energy conversion efficiency of heat source tower heat pumps, The historical status data can also include temperature data (such as annual average temperature, minimum temperature in winter, maximum temperature in summer, temperature fluctuation range, etc.), humidity data (such as annual average relative humidity, seasonal humidity changes), precipitation (such as annual rainfall, precipitation distribution, frequency of extreme precipitation events), sunshine hours (such as annual average sunshine hours, changes in sunshine intensity), wind speed (such as annual average wind speed, seasonal changes in wind speed), wind direction (such as dominant wind direction, wind direction changes), geographical factors (such as altitude, terrain features), extreme weather events (such as cold wave frequency, heat wave duration, natural disaster frequency), etc. These data can help the neural network model understand the climate characteristics in a specific area, identify energy consumption patterns in different regions and seasons, optimize energy utilization, increase the universality of the heat source tower heat pump applicability evaluation model, reduce the peak load of the power system, and alleviate the system load pressure; at the same time, more efficiently use electricity in the heating and cooling process, reduce the load fluctuation of the power system, reduce carbon emissions and energy consumption, thereby optimizing the applicability evaluation model of the heat source tower heat pump system to adapt to local climate conditions and improve energy conversion efficiency. In addition, these data can also help determine whether the heat source tower heat pump system is suitable for a specific geographical area and how to adjust the system parameters to achieve optimal performance, providing a theoretical basis for the construction and operation of the heat source tower heat pump.

[0100] In view of the problem that the evaluation indicators of the existing heat source tower heat pump applicability evaluation model are not comprehensive enough, the key indicators in the comprehensive heat source tower heat pump applicability evaluation index system constructed by the present invention may include: system performance indicators, energy-saving and environmental protection performance indicators and economic indicators; wherein the system performance indicators are used to characterize the operating performance efficiency of the heat source tower heat pump system under winter heating and summer cooling conditions. For example, the system performance indicators may include: the cooling capacity of the heat source tower heat pump, the heating capacity of the heat source tower heat pump, the winter heating performance coefficient of the heat source tower heat pump host, the summer cooling performance coefficient of the heat source tower heat pump host and the system comprehensive performance coefficient Energy-saving and environmental protection performance indicators are used to characterize the equipment power consumption and energy consumption of the heat source tower heat pump system under the premise of meeting the requirements of refrigeration and heating, as well as the reduction in carbon emission costs. For example, energy-saving and environmental protection performance indicators may include: carbon emission costs and system energy consumption coefficients; economic indicators are used to characterize the economic costs of energy consumption and maintenance of the heat source tower heat pump system during operation. For example, economic indicators may include: investment costs and operating costs; In the present invention, the applicability of the heat source tower heat pump is quantitatively analyzed based on the key indicators in the comprehensive heat source tower heat pump applicability evaluation index system. The specific analysis process may include:

[0101] 1) System performance index analysis: Since the heat source tower heat pump is affected by regional differences and climate environment, its operating efficiency will change. The heat pump cooling / heating capacity, the host winter heating and summer cooling performance coefficient (Coefficient of Performance, COP) and the system comprehensive performance coefficient (Seasonal Energy Efficiency Ratio, SEER) are used to characterize the operating performance of the heat source tower heat pump system. The heating capacity and cooling capacity of the heat source tower heat pump can be expressed by the following formula:

[0102] Q=Cp w m w (t w,i -t w,o );

[0103] Among them, Cp w represents the specific heat capacity at constant pressure, in units of kJ / (kg·K); when Q represents the heating capacity of the heat source tower heat pump, m w Indicates the air mass flow rate under heating conditions, t w,i Indicates the water temperature at the outlet side of the unit under heating conditions, t w,o represents the water temperature at the water inlet side of the unit under heating conditions; when Q represents the cooling capacity of the heat pump in the heat source tower, m w It represents the air mass flow rate under refrigeration conditions; t w,i Indicates the water temperature at the outlet side of the unit under refrigeration conditions, t w,o Indicates the water temperature at the water inlet side of the unit under refrigeration conditions.

[0104] The coefficient of performance (COP) refers to the heating or cooling capacity per unit power consumption of the heating and cooling units under operating conditions. The calculation formula can be shown as follows:

[0105]

[0106] Wherein, P represents the host input power (unit: W); Q represents the cooling capacity or heating capacity of the heat source tower heat pump;

[0107] The system comprehensive performance coefficient SEER refers to the ratio of the total winter heating load to the total winter power consumption or the ratio of the total summer cooling load to the total summer power consumption. The calculation formula can be shown as follows:

[0108]

[0109] Among them, Q c Indicates the heating or cooling capacity of the heat source tower heat pump in summer (unit: kW); W comp is the power consumption of the compressor (unit: kW); W pump is the power consumption of the heat pump side of the heat source tower (unit: kW); W fan It is the power consumption of the fan on the heat pump side of the heat source tower (unit: kW).

[0110] 2) Energy-saving and environmental protection performance index analysis: In this embodiment, the carbon emission cost and the system energy consumption coefficient (such as the energy quality coefficient) are used to characterize the energy-saving and environmental protection of the system; wherein, the calculation formula corresponding to the carbon emission cost can be as follows:

[0111]

[0112] Among them, E rs represents the carbon emission cost, which is the sum of carbon dioxide emissions generated by the combustion of fossil fuels by various power generation facilities in the integrated energy system; is the consumption of the ith fossil fuel; is the as-received carbon content of the ith fossil fuel; is the carbon oxidation rate of the ith fossil fuel; 44 / 12 is the relative molecular mass ratio of carbon dioxide to carbon; n is the total number of fossil fuels;

[0113] The energy quality coefficient is used to express the ratio of the work that different energy sources can do to their total energy at the current technological level. The calculation formula can be shown as follows:

[0114]

[0115] Among them, λ is the energy quality coefficient of a certain form of energy; W0 is the part of the total energy of this form of energy that can be converted into work (unit: W); N0 is the total energy of this form of energy (unit: W).

[0116] Based on the energy quality coefficients of various forms of energy, the system energy consumption coefficient of the heat source tower heat pump can be obtained, and its calculation formula can be shown as follows:

[0117]

[0118] Where ECC represents the system energy consumption coefficient of the heat source tower heat pump; E represents the electric energy consumed by the air conditioning system (unit: kW·h); E a is the energy consumption of the ath energy form in the air conditioning system (unit: kW·h); a is the energy quality coefficient of the ath energy form; X is the cooling or heating demand of the building (kW·h).

[0119] 3) Economic index analysis: Investment cost and operating cost are used to characterize the energy conservation and environmental protection of the system; investment cost can include equipment procurement cost, installation cost, land and construction cost, etc. The level of investment cost directly affects the economy of the system. Generally speaking, lower investment cost means a shorter investment recovery period and a more attractive economic return. Due to the wide range of values ​​of the design parameter matrix, it cannot be guaranteed that the systems corresponding to all matrices can meet the hourly load of the building, especially under some extreme weather conditions. Therefore, in this embodiment, auxiliary chillers and auxiliary electric heating are introduced to ensure that the building load can be met during the simulation. The additional energy consumption generated is included in the calculation of various indicators as a penalty factor under the corresponding design parameter scheme. At the same time, the use of chillers under summer conditions will also generate additional initial investment (for example, the cost of chillers, which can be calculated using IC chiller =Indicates) as an economic penalty included in the initial investment of the system (for example, using IC sys The cost of heat pump, heat source tower, solution pump, operating solution and water are all included in the expression as follows:

[0120] IC sys =IC HP +IC HT +IC chiller +IC re +IC s +IC water ;

[0121] Among them, IC sys is the initial investment of the system (10,000 yuan); IC HP is the cost of the heat pump host (10,000 yuan); IC HT is the cost of the heat source tower (10,000 yuan); IC chilleris the cost of the chiller (10,000 yuan); IC re is the solution pump cost (10,000 yuan); IC S is the cost of operating solution (10,000 yuan); IC water is the water cost (10,000 yuan);

[0122] Operating cost is an important part of the economic analysis of the new heat source tower heat pump in long-term operation. It can include energy consumption costs, maintenance costs, and operation and management costs. Through detailed cost analysis, the operating cost of the system can be evaluated, and the main costs can be optimized to reduce the overall operating cost of the system. The operating cost (OC) is calculated based on energy costs and energy consumption. The specific calculation formula can be as follows:

[0123] OC sys =P elec +∫(W comp +W ac +W ah +W re +W fan +W pump )dt;

[0124] Among them, OC sys is the total operating cost of the system (unit: 10,000 yuan); P elec is the energy cost (unit: 10,000 yuan); W ac is the power consumption of the auxiliary cooling device within the time interval dt (unit: kW); W ah is the power consumption of the auxiliary heating device within the time interval dt (unit: kW); W re is the power consumption of the regeneration device within the time interval dt (unit: kW); W comp is the power consumption of the compressor in the time interval dt (unit: kW); W fan is the fan power consumption in the time interval dt (unit: kW); W pump is the power consumption of the pump during the time interval dt (in kW);

[0125] In this embodiment, the present invention establishes a universal applicability evaluation index system centered around evaluation indicators such as system performance, energy saving, and economy based on the key indicators of comprehensive evaluation of the applicability of the heat source tower heat pump, considers the characteristic constraints of the heat source tower heat pump itself, establishes a heat source tower heat pump applicability comprehensive evaluation index system and an index calculation method, analyzes the system performance, energy saving, environmental protection, and economy of the heat source tower heat pump, and obtains the system initial data set based on the calculation of each secondary indicator host performance, system comprehensive performance, energy consumption coefficient of the heat source tower heat pump system, carbon emission cost, initial investment cost, and operating cost, which is conducive to more accurate evaluation of the heat source tower heat pump The applicability of the technology in different scenarios can improve energy utilization efficiency and promote clean energy transformation, provide a comprehensive and accurate evaluation framework for the application of heat source tower heat pump technology, solve the rationality of the evaluation index weights, the evaluation method is affected by subjective factors, and the evaluation index is not comprehensive enough, so as to quantitatively evaluate its applicability and performance in different environments and application scenarios; at the same time, provide development suggestions and theoretical support for decision makers in the construction and planning of heat source tower heat pump systems, so as to promote the widespread application of heat source tower heat pumps, reduce energy waste, reduce environmental impact, assist decision makers in making more informed investment decisions, and promote the application of green energy technology.

[0126] The index values ​​of the above historical state data under the constructed applicability evaluation index system are used as the input data of the training data; the actual applicability evaluation results corresponding to the index values ​​are used as the output data of the training data, and the neural network model is trained using the particle swarm algorithm to obtain the applicability evaluation model; wherein, the above historical state data refers to the operation data of the heat source tower heat pump being homogenized and mapped to a specified range to remove the dimensions and dimensional units of data of different dimensions; for example, the expression of the homogenization function is:

[0127]

[0128] Among them, x max Indicates the maximum value in the historical status data of the heat source tower heat pump; x min represents the minimum value in the historical state data of the heat source tower heat pump; x represents the training data; x new Represents the historical state data of the heat source tower heat pump after homogenization; the homogenized data is used as the input data in the training and test data sets, the sampling data set needs to include input data x and output data y, and the data set is divided into a training set and a test set in a preset ratio (for example, a ratio of 4:1), and the training is performed with the real applicability evaluation result as the expected output. For example, the above applicability evaluation results are divided into four levels, namely unsuitable, suitable, relatively suitable and very suitable, to illustrate the real adaptability of the heat source tower heat pump.

[0129] In one implementation, the process of training the neural network model using the particle swarm algorithm to obtain the suitability evaluation model may include:

[0130] Based on input data, output data, randomly generated initial weights and initial thresholds, the neural network model is trained using the particle swarm algorithm to obtain the optimal weights and optimal thresholds;

[0131] According to the optimal weight and the optimal threshold, a suitability evaluation model is obtained; in this implementation, the neural network model is trained using a particle swarm algorithm, which can search for a global optimal solution or a weight and threshold combination close to the optimal solution, thereby improving the prediction accuracy of the neural network model. By optimizing the weights and thresholds, the neural network model can perform better on unseen data sets, reducing the complexity of manually adjusting hyperparameters, thereby improving the generalization ability of the model.

[0132] However, if the traditional particle swarm algorithm is used to train the neural network model and build a suitability evaluation model, it is easy to fall into local extreme values ​​and premature convergence and stagnation. Therefore, the particle swarm algorithm needs to be optimized. Specifically:

[0133] In one implementation, the process of using the particle swarm algorithm to train the neural network model to obtain the optimal weight and the optimal threshold may include:

[0134] Initialize the parameters of the particle swarm algorithm, including the number of particles in the particle swarm, particle positions and number of iterations, as well as initialize individual extreme values ​​and global optimal extreme values. The initial position of each particle corresponds to a set of initial weights and initial thresholds of the neural network model.

[0135] Input the input data of the training data into the neural network model to obtain the prediction data corresponding to the input data;

[0136] Calculate the fitness function value of each particle based on the square of the difference between the predicted data and the output data;

[0137] Based on the fitness function value of each particle, the particle swarm algorithm is used to iteratively update the individual extreme value and the global optimal extreme value until the updated individual extreme value and the global optimal extreme value reach the convergence condition. At this time, the preset inertia weight update expression is used to update the initial weight and the initial threshold, and the updated initial weight is determined as the optimal weight, and the updated initial threshold is determined as the optimal threshold; in this implementation method, in the process of iteratively updating the individual extreme value and the global optimal extreme value using the particle swarm algorithm, if the updated individual extreme value and the global optimal extreme value do not reach the convergence condition, the particle velocity update expression is used to update the particle position until the updated individual extreme value and the global optimal extreme value converge. In the particle swarm algorithm, particles iteratively update their own flight speed through individual extreme values ​​and global optimal extreme values. However, this process may bring some technical problems, especially when using the default learning factor. If the learning factor is too large, the particles may quickly converge to the current optimal solution, but this often leads to premature convergence. If the learning factor is too small, the update speed of the particles will become very slow, resulting in slow convergence, or even failure to find a satisfactory solution within a limited time. Inappropriate learning factors may lead to insufficient exploration of particles in the search space, especially in high-dimensional or complex optimization problems, particles may miss the global optimal solution. Therefore, in this implementation, the particle speed update formula is improved, and the learning factor is dynamically adjusted according to the particle position, speed and number of iterations, so that the algorithm focuses on global search in the early stage of the search and local search in the later stage. The performance of the particle swarm algorithm can be effectively improved, so that it can converge quickly during the optimization process and avoid premature convergence and local optimal problems. Specifically, the particle speed update expression can be as follows:

[0138]

[0139] in, represents the flying speed of particle i in the dth dimension at the t+1th iteration; ω max Represents the maximum value of inertia weight; ω min represents the minimum value of the inertia weight; t represents the current number of iterations; t max represents the maximum number of iterations; e is a natural constant; v′ id represents the iteration speed of particle i in the dth dimension; R1 represents the fuzzy set relationship matrix of the first learning factor; R2 represents the fuzzy set relationship matrix of the second learning factor; represents the individual extreme value of particle i at the tth iteration; represents the global optimal extreme value of particle i at the tth iteration; represents the position of particle i in the dth dimension at the tth iteration; c1(t) represents the first learning factor corresponding to the tth iteration; c2(t) represents the second learning factor corresponding to the tth iteration;

[0140] In the above formula, the expression corresponding to the first learning factor c1(t) is as follows:

[0141]

[0142] In the above formula, the expression corresponding to the second learning factor c2(t) is as follows:

[0143]

[0144] Among them, c 1e represents the initial value of the first learning factor; c 1s represents the termination value of the first learning factor; c 2e represents the initial value of the second learning factor; c 2s represents the termination value of the second learning factor; r1 represents the first random number; r2 represents the second random number;

[0145] In this example, by improving the learning factor in the particle swarm algorithm, it is possible to solve the problem that when faced with complex problems, the search ability of the particle swarm algorithm will fall into local optimality as the learning factor c1 decreases and c2 increases, resulting in slower algorithm convergence and a smaller possibility of obtaining the optimal solution.

[0146] In one implementation, the process of updating the initial weight and the initial threshold using the preset inertia weight update expression may include:

[0147] The inertia weight of the particle is updated using a preset inertia weight update expression to obtain an updated inertia weight;

[0148] Determine whether the maximum number of iterations has been reached. If so, update the initial weight and initial threshold based on the updated inertia weight; if not, update the speed of each particle based on the updated inertia weight using the above particle speed update expression, and obtain the new position of each particle based on the updated particle speed, and calculate the fitness function value for the new position of each particle to determine the new individual extreme value and the global optimal extreme value.

[0149] In the particle swarm algorithm, inertia weight is generally used to balance the local and global search capabilities of particles. In order to ensure that particles have good local search capabilities in the early stage of iteration and can converge quickly in the later stage, a linear decreasing strategy is usually used to dynamically adjust the inertia weight (for example, represented by ω). Since the search process of particles in the particle swarm algorithm is nonlinear and complex, in order to better reflect the actual search process of particles, the present invention uses a nonlinear function to construct an inertia weight update expression and dynamically adjusts its inertia weight. Specifically, the inertia weight update expression can be as follows:

[0150]

[0151] Among them, ω(t) represents the inertia weight at the tth iteration; ωma x Represents the maximum value of inertia weight; ω min represents the minimum value of the inertia weight; t represents the current number of iterations; t max Represents the maximum number of iterations; e is a natural constant.

[0152] For example, the update expression of the above initial weight can be as follows:

[0153] w(t+1)=w(t)+η(P best (t)-w(t))+η(G best (t)-w(t));

[0154] For example, the update expression of the above initial threshold can be as follows:

[0155]

[0156] Wherein, w(t+1) represents the initial weight of the neural network model at the t+1th iteration; w(t) represents the initial weight of the neural network model at the tth iteration; η represents the learning rate of the neural network model; P best (t) represents the individual extreme value at the tth iteration; G best (t) represents the global optimal extreme value at the t-th iteration; b(t+1) represents the initial threshold at the t+1-th iteration; b(t) represents the initial threshold at the t-th iteration; Represents the output data at the tth iteration; Represents the predicted data corresponding to the input data at the tth iteration.

[0157] In this implementation, the inertia weight of the particle is updated by using a preset inertia weight update expression, and the initial weight and the initial threshold are updated according to the updated inertia weight. In this way, a larger inertia weight can be maintained in the early stage of the search to enhance the global search capability, while the inertia weight can be reduced in the later stage of the search to enhance the local search capability, and the speed of the particle is changed in time according to the current search state, thereby accelerating the convergence speed. The performance of the particle swarm algorithm (PSO) in optimizing the neural network model can be significantly improved, thereby bringing better technical effects, especially when dealing with high-dimensional or complex problems, it can effectively utilize computing resources and improve search efficiency.

[0158] In one implementation, the process of updating the initial weight and the initial threshold based on the updated inertia weight may include:

[0159] According to the updated inertia weight, the learning rate of the neural network model is obtained;

[0160] According to the learning rate of the neural network model, the initial weights and initial thresholds are updated through the back-propagation algorithm.

[0161] In this implementation, by adjusting the learning rate based on the updated inertia weight and updating the initial weight and initial threshold through the back propagation algorithm, the training effect of the neural network model can be significantly improved, and a better weight and threshold configuration can be obtained, thereby improving the prediction accuracy and generalization ability of the model. In addition, the method of dynamically adjusting the learning rate can support optimization in a parallel computing environment, so that the algorithm can run more efficiently in a distributed computing environment. Therefore, the method of the present invention can improve the convergence speed, avoid premature convergence, enhance the generalization ability and robustness of the model, and simplify the hyperparameter tuning process, thereby bringing better technical effects.

[0162] For example, the neural network model can be a back propagation (BP) neural network model. In this example, when the BP neural network is trained using the particle swarm algorithm, it is necessary to determine the topological structure of the BP neural network model and initialize the weights and thresholds of the network. For example, the network structure of the BP neural network model is divided into an input layer, an intermediate layer, and an output layer. After determining the network structure of the model, it is necessary to further determine the optimal weights and thresholds in the model so that the model network can predict accurately and minimize the error. The flowchart of training the BP neural network model is as follows: Figure 2 As shown, the specific steps can be as follows:

[0163] (1) Initialize the particle swarm, set the number of particles in the swarm to N, and randomly initialize a group of particles. The initial parameters of the particles include particle position, velocity, individual extreme value, and global extreme value. In the PSO algorithm, the position and iteration speed of the i-th particle distributed in the D-dimensional space are represented as X i and V i , the position of each particle is regarded as the weight and threshold of the neural network, and the speed represents the adjustment of the corresponding parameters. The individual extreme value (individual optimal position) of the initial particle is P 0best , the initial global extremum (global optimal position) is G 0best .

[0164] (2) Calculate the particle output value. Input the data in the training set into the network model to obtain the predicted output data. Use the improved inertia weight ω(t) of the PSO algorithm to calculate the output values ​​of the hidden layer and the output layer. The jth output h of the hidden layer j and the output layer i-th output y i The expressions are:

[0165]

[0166] Where b1 and b2 are the thresholds of the input layer and the hidden layer respectively; i = 1…n; n represents the total number of outputs of the output layer; j = 1…J; J represents the total number of outputs of the hidden layer; ω ji (t) represents the inertia weight corresponding to the j-th output of the hidden layer and the i-th output of the output layer; x i Represents the independent variable corresponding to the i-th output of the output layer; ω kj (t) represents the inertia weight corresponding to the k-th output of the hidden layer and the j-th output of the output layer; h j represents the independent variable corresponding to the j-th output of the output layer; f is the activation function, and the expression is:

[0167]

[0168] Among them, x is the independent variable of the activation function;

[0169] (3) Determine the value of the particle fitness function and update the individual extreme value of the particle and the global optimal extreme value. The fitness function value of each particle is calculated using the following fitness function:

[0170]

[0171] In the formula, represents the output value of the ith particle (given in the training set), represents the predicted output value of the ith particle; Fitness represents the fitness function value of the ith particle; the fitness function values ​​of all particles in the population are calculated by the above formula, and the fitness function value of the optimal position of the initial particle is determined according to the optimal position of the initial particle. If the fitness function value of the optimal position of the particle is less than the fitness function value of the optimal position experienced by the particle, the current position is used to replace the individual historical optimal position P best According to the fitness function value of each particle, the fitness function value of the optimal position of all particles in the updated population is compared with the fitness function value of the optimal position experienced by the population. If the fitness function value of the former is smaller, the global optimal position G is updated. best .

[0172] (4) Update particle speed and position. Based on the obtained individual extreme value of the particle and the global extreme value of the particle group (the first cycle is the initial individual extreme value and the global optimal extreme value), the flight speed of each particle is updated using the improved particle flight speed update formula. The position of each particle is updated according to the following particle flight speed update formula:

[0173] X id t+1 =X id i +V id t+1 ;

[0174] in, represents the position of particle i in the dth dimension at the tth iteration; represents the flying speed of particle i in the dth dimension at the t+1th iteration;

[0175] (5) Update model weights and thresholds. Based on the updated individual extreme values ​​and global optimal extreme values, the updated inertia weights are used to calculate the model learning rate η, and the model weights and thresholds are updated. The specific update formula is as follows:

[0176] w(t+1)=w(t)+η(P best (t)-w(t))+η(G best (t)-w(t));

[0177]

[0178] Wherein, w(t+1) represents the initial weight of the neural network model at the t+1th iteration; w(t) represents the initial weight of the neural network model at the tth iteration; η represents the learning rate of the neural network model (η>0); P best (t) represents the individual extreme value at the tth iteration; G best (t) represents the global optimal extreme value at the t-th iteration; b(t+1) represents the initial threshold at the t+1-th iteration; b(t) represents the initial threshold at the t-th iteration; Represents the output data at the tth iteration; Represents the predicted data corresponding to the input data at the tth iteration.

[0179] (6) Repeat steps (2), (3), (4) and (5) until the model training reaches the maximum number of iterations. After the algorithm terminates, record the optimal weight and threshold at this time, and output the global optimal position value.

[0180] After the model training is completed, the model is tested: the optimal weights and thresholds obtained after the iteration are assigned to the BP neural network model, which is the optimal suitability evaluation model. The test set input data x1 is input into the optimal suitability evaluation model to obtain the model output value y0, and the mean square error function is used to calculate the output error of the suitability evaluation model:

[0181]

[0182] Where F represents the output error of the suitability evaluation model; represents the model output value of the i-th particle in the test set; represents the actual model value of the i-th particle in the test set; the number of particles in the population is N; after obtaining the output error of the optimal model, calculate the output error of the model before optimization, compare the output errors of the two, and judge the performance of the optimized model. The smaller the error, the better the performance of the optimized model. For example, the applicability evaluation weight values ​​that may be obtained based on the applicability evaluation model are shown in Table 1:

[0183] Table 1 Weights of comprehensive evaluation indicators for heat source tower heat pump applicability

[0184]

[0185]

[0186] The network architecture of the heat source tower heat pump suitability evaluation model based on particle swarm algorithm includes modules such as input data, data preprocessing, neural network model, data processing, and data output. The neural network model includes three parts: BP neural network, error evaluation, and improved particle swarm algorithm. The input data is the index values ​​of the basic indicators in the input suitability evaluation index system; data preprocessing is to make each data have a unified standard and definition and reduce the data structure difference; the neural network model is to train the network by inputting enough samples, and adaptively learn the network weight coefficient to obtain preliminary results. The improved particle swarm algorithm is used to solve the error function of the neural network model (for example, BP neural network) to solve the optimal network parameters of the neural network; data processing is used to measure the performance of the model; output data outputs the suitability evaluation results of the heat source tower heat pump. Compared with other neural network models, BP neural network has strong nonlinear mapping ability and can handle the complex input-output relationship in the heat source tower heat pump system well. BP neural network can automatically adjust weights and biases through training and has the ability of self-learning, which means that as data accumulates, the network can continuously optimize its own performance and adapt to new input and output modes. In addition, BP neural network has a certain fault tolerance. Even if some data is lost or there is noise, the network can still give a relatively accurate prediction result. Therefore, in practical applications, using particle swarm algorithm to train BP neural network can make the obtained suitability evaluation model more stable and accurate. Multiply the obtained index value by the weight value to get the suitability evaluation score. The specific output adaptability evaluation result is expressed as a score:

[0187]

[0188] Where: is the index value after normalization of the i-th data, is the weight of the ith indicator, C is the suitability evaluation result. For example, the value of C can be defined as unsuitable when it is between 0 and 0.6, relatively suitable when it is between 0.6 and 0.8, suitable when it is between 0.8 and 0.9, and very suitable when it is between 0.9 and 1.

[0189] In summary, the present invention aims at the problem that the existing applicability evaluation model is affected by subjective factors, resulting in a low accuracy of model prediction, and proposes a method for constructing a heat source tower heat pump applicability evaluation model. From the perspective of the heat source tower heat pump's own characteristic constraints, the influence of regional and climate differences on the performance of the heat source tower heat pump system is analyzed to construct a heat source tower heat pump applicability evaluation system; and an improved particle swarm algorithm with improved inertia weight, particle update speed and learning factor is proposed. The improved particle swarm algorithm is used to establish a suitability evaluation model based on the applicability evaluation index system, the particle swarm algorithm and the neural network model, so as to more accurately evaluate the applicability of the heat source tower heat pump system. Through the method proposed by the present invention, the particle swarm optimization and the neural network are combined to realize the combination of advanced computer algorithms and energy technology evaluation, which can not only reduce the subjectivity of human judgment, improve the fitting accuracy of the neural network, and increase the accuracy of model prediction, but also improve the comprehensive evaluation ability of the heat source tower heat pump technology, more accurately simulate and predict system performance, help reduce the trial and error cost, and improve the success rate of technology application, thereby promoting the development and application of clean energy technology and promoting the utilization of sustainable energy.

[0190] Embodiment 2:

[0191] Based on the same inventive concept, the present invention also provides a heat source tower heat pump applicability evaluation model construction system, the structural composition diagram is as follows: Figure 3 As shown, including:

[0192] A data acquisition module is used to obtain the historical status data of the heat source tower heat pump, the index values ​​of the historical status data under the preset applicability evaluation index system, and the real applicability evaluation results corresponding to the index values;

[0193] A training input module is used to use the values ​​of each indicator of the historical state data under the applicability evaluation indicator system as input data of the training data;

[0194] A training output module is used to output the actual applicability evaluation results corresponding to each indicator value as the output data of the training data;

[0195] The model building module is used to train the neural network model based on input data and output data using the particle swarm algorithm to obtain a suitability evaluation model.

[0196] For example, the above neural network model is a back propagation neural network model.

[0197] In one implementation, the above-mentioned model building module may include:

[0198] The model training submodule is used to train the neural network model based on the input data, output data, randomly generated initial weights and initial thresholds using the particle swarm algorithm to obtain the optimal weights and optimal thresholds;

[0199] The parameter adjustment submodule is used to obtain a suitability evaluation model according to an optimal weight and an optimal threshold.

[0200] In this implementation, the above-mentioned model training submodule may include:

[0201] Parameter initialization unit, used to initialize the parameters of the particle swarm algorithm, including the number of particles in the particle swarm, particle positions and number of iterations, as well as initializing individual extreme values ​​and global optimal extreme values. The initial position of each particle corresponds to a set of initial weights and initial thresholds of the neural network model.

[0202] A model output unit, used to input the input data of the training data into the neural network model to obtain prediction data corresponding to the input data;

[0203] A fitness calculation unit, used to calculate the fitness function value of each particle based on the square of the difference between the predicted data and the output data;

[0204] The parameter updating unit is used to iteratively update the individual extreme value and the global optimal extreme value based on the fitness function value of each particle by using the particle swarm algorithm until the updated individual extreme value and the global optimal extreme value reach the convergence condition. At this time, the preset inertia weight update expression is used to update the initial weight and the initial threshold, and the updated initial weight is determined as the optimal weight, and the updated initial threshold is determined as the optimal threshold.

[0205] In this implementation, the parameter updating unit may include:

[0206] The weight updating subunit is used to update the inertia weight of the particle using a preset inertia weight updating expression to obtain an updated inertia weight;

[0207] The parameter determination subunit is used to determine whether the maximum number of iterations has been reached. If so, the initial weight and initial threshold are updated based on the updated inertia weight.

[0208] In this implementation, the above-mentioned parameter determination subunit may include:

[0209] A learning rate update component is used to obtain the learning rate of the neural network model according to the updated inertia weight;

[0210] The optimal parameter updating component is used to update the initial weights and initial thresholds according to the learning rate of the neural network model through the back propagation algorithm.

[0211] For example, the above inertia weight update expression can be as follows:

[0212]

[0213] Among them, ω(t) represents the inertia weight at the tth iteration; ω max Represents the maximum value of inertia weight; ω min represents the minimum value of the inertia weight; t represents the current number of iterations; t max Represents the maximum number of iterations; e is a natural constant.

[0214] For example, the update expression of the above initial weight can be as follows:

[0215] w(t+1)=w(t)+η(P best (t)-w(t))+η(G best (t)-w(t));

[0216] For example, the update expression of the above initial threshold can be as follows:

[0217]

[0218] Wherein, w(t+1) represents the initial weight of the neural network model at the t+1th iteration; w(t) represents the initial weight of the neural network model at the tth iteration; η represents the learning rate of the neural network model; P best (t) represents the individual extreme value at the tth iteration; G best (t) represents the global optimal extreme value at the t-th iteration; b(t+1) represents the initial threshold at the t+1-th iteration; b(t) represents the initial threshold at the t-th iteration; Represents the output data at the tth iteration; Represents the predicted data corresponding to the input data at the tth iteration.

[0219] Embodiment 3:

[0220] Based on the same concept, the present invention also provides a heat source tower heat pump applicability evaluation method, the flow diagram is as follows Figure 4 As shown, including:

[0221] Step S1: Obtaining status data of the heat source tower heat pump;

[0222] Step S2: Based on the state data, the suitability evaluation model is used to evaluate the suitability of the heat source tower heat pump to obtain a suitability evaluation result of the heat source tower heat pump;

[0223] The suitability evaluation model is constructed using the heat source tower heat pump suitability evaluation model construction method as described above.

[0224] For example, the above-mentioned comprehensive evaluation index system of applicability may include: system performance index, energy-saving and environmental protection performance index and economic index;

[0225] For example, the above-mentioned system performance indicators may include one or more of the following: cooling capacity of the heat source tower heat pump, heating capacity of the heat source tower heat pump, winter heating performance coefficient of the heat source tower heat pump host, summer cooling performance coefficient of the heat source tower heat pump host and system comprehensive performance coefficient;

[0226] For example, the energy-saving and environmental-friendly indexes mentioned above may include one or more of the following: carbon emission cost and system energy consumption coefficient;

[0227] For example, the above-mentioned economic indicators may include one or more of the following: investment cost and operating cost;

[0228] By way of example, the above-mentioned status data may include one or more of the following: heating capacity, cooling capacity, air mass flow rate, water temperature on the outlet side of the unit, water temperature on the inlet side of the unit, host input power, compressor power consumption, tower pump side power consumption, tower side fan power consumption, fossil fuel consumption, basic carbon content of fossil fuels, carbon oxidation rate of fossil fuels, relative molecular mass ratio of carbon dioxide to carbon, power consumption of air-conditioning system, energy consumption, cooling demand of associated buildings, heating demand of associated buildings, equipment procurement cost, installation cost, land cost, construction cost, host cost, chiller cost, solution pump cost, operating solution cost and water cost.

[0229] The energy conversion efficiency of the heat source tower heat pump is related to many factors, such as heat source temperature, compressor efficiency, external climate and environmental conditions, regional differences, and the design and type of the heat pump. As a result, when evaluating the applicability of the heat source tower heat pump, the existing heat source tower heat pump applicability evaluation model has weak universality and the evaluation indicators are not comprehensive enough. Therefore, when constructing the applicability evaluation model of the heat source tower heat pump, the historical status data used in the present invention takes into account many factors, such as: heating capacity, cooling capacity, air mass flow, water temperature on the water outlet side of the unit, water temperature on the water inlet side of the unit, main The machine input power, compressor power consumption, tower pump side power consumption, tower side fan power consumption, fossil fuel consumption, basic carbon content of fossil fuels, carbon oxidation rate of fossil fuels, relative molecular mass ratio of carbon dioxide to carbon, power consumption of air conditioning system, energy consumption, cooling demand of related buildings, heating demand of related buildings, equipment procurement costs, installation costs, land costs, construction costs, host costs, chiller costs, solution pump costs, operating solution costs and water costs, and considering that climate and regional differences have a great impact on the energy conversion efficiency of heat source tower heat pumps, Historical status data can also include temperature data (such as annual average temperature, minimum temperature in winter, maximum temperature in summer, temperature fluctuation range, etc.), humidity data (such as annual average relative humidity, seasonal humidity changes), precipitation (such as annual rainfall, precipitation distribution, frequency of extreme precipitation events), sunshine hours (such as annual average sunshine hours, changes in sunshine intensity), wind speed (such as annual average wind speed, seasonal changes in wind speed), wind direction (such as dominant wind direction, wind direction changes), geographical factors (such as altitude, terrain features), extreme weather events (such as cold wave frequency, heat wave duration, natural disaster frequency), etc. These data can help the neural network model understand the climate characteristics in a specific area, identify energy usage patterns in different regions and seasons, optimize energy utilization, increase the universality of the heat source tower heat pump applicability evaluation model, reduce the peak load of the power system, and alleviate the system load pressure; at the same time, more efficiently use electricity in the heating and cooling process, reduce load fluctuations in the power system, reduce carbon emissions and energy consumption, thereby optimizing the applicability evaluation model of the heat source tower heat pump system to adapt to local climate conditions and improve energy conversion efficiency. In addition, these data can also help determine whether the heat pump system is suitable for a specific geographical area and how to adjust the system parameters to achieve optimal performance, providing a theoretical basis for the construction and operation of heat source tower heat pumps.

[0230] The present invention aims at the problem that the applicability evaluation of the existing heat source tower heat pump is affected by subjective factors, resulting in low accuracy of the applicability evaluation result. When evaluating the applicability of the heat source tower heat pump, Figure 5As shown, the suitability evaluation result is obtained by using a suitability evaluation model constructed based on a suitability evaluation index system, a particle swarm algorithm and a neural network model (such as a BPNN neural network model), which can reduce the subjectivity of human judgment and increase the accuracy of the suitability evaluation of the heat source tower heat pump.

[0231] Embodiment 4:

[0232] Based on the same inventive concept, the present invention also provides a heat source tower heat pump applicability evaluation system, the structural composition diagram is as follows: Figure 6 As shown, including:

[0233] The operation data acquisition module is used to obtain the status data of the heat pump of the heat source tower;

[0234] The applicability evaluation module is used to evaluate the applicability of the heat source tower heat pump according to the state data using the applicability evaluation model to obtain the applicability evaluation result of the heat source tower heat pump;

[0235] The suitability evaluation model is constructed using the heat source tower heat pump suitability evaluation model construction method as described above.

[0236] For example, the above-mentioned comprehensive evaluation index system of applicability may include: system performance index, energy-saving and environmental protection performance index and economic index;

[0237] For example, the above-mentioned system performance indicators may include one or more of the following: cooling capacity of the heat source tower heat pump, heating capacity of the heat source tower heat pump, winter heating performance coefficient of the heat source tower heat pump host, summer cooling performance coefficient of the heat source tower heat pump host and system comprehensive performance coefficient;

[0238] For example, the energy-saving and environmental-friendly indexes mentioned above may include one or more of the following: carbon emission cost and system energy consumption coefficient;

[0239] For example, the above-mentioned economic indicators may include one or more of the following: investment cost and operating cost;

[0240] By way of example, the above-mentioned status data may include one or more of the following: heating capacity, cooling capacity, air mass flow rate, water temperature on the outlet side of the unit, water temperature on the inlet side of the unit, host input power, compressor power consumption, tower pump side power consumption, tower side fan power consumption, fossil fuel consumption, basic carbon content of fossil fuels, carbon oxidation rate of fossil fuels, relative molecular mass ratio of carbon dioxide to carbon, power consumption of air-conditioning system, energy consumption, cooling demand of associated buildings, heating demand of associated buildings, equipment procurement cost, installation cost, land cost, construction cost, host cost, chiller cost, solution pump cost, operating solution cost and water cost.

[0241] Embodiment 5:

[0242] like Figure 7 As shown, the present invention also provides an electronic device, which may be a computer device, a single-chip device, an intelligent mobile device, etc. The electronic device in this embodiment may include a processor, a memory, a transceiver component, etc. The memory, the processor, and the transceiver component are connected via a bus; the memory may be used to store an execution program, and an exemplary execution program may include instructions; the processor is used to execute the instructions stored in the memory. The memory may also be used to store data, which may be called and / or modified when the instructions are executed.

[0243] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in a storage medium to implement corresponding method flows or corresponding functions, so as to implement a method for constructing a heat source tower heat pump suitability evaluation model or a step of a heat source tower heat pump suitability evaluation method in the above-mentioned embodiment.

[0244] Embodiment 6:

[0245] Based on the same inventive concept, the present invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory), which is a memory device in an electronic device for storing programs and data. It is understandable that the storage medium here can include both built-in storage media in electronic devices and, of course, extended storage media supported by electronic devices. The storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by a processor are also stored in the storage space, and these instructions can be one or more execution programs (including program codes). It should be noted that the storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor loads and executes one or more instructions stored in the storage medium, which can implement the steps of a heat source tower heat pump applicability evaluation model construction method or a heat source tower heat pump applicability evaluation method in the above-mentioned embodiment.

[0246] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0247] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0248] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0249] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0250] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit its protection scope. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that after reading the present invention, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the application, but these changes, modifications or equivalent substitutions are all within the protection scope of the claims to be approved.

Claims

1. A method for constructing a heat source tower heat pump suitability evaluation model, characterized in that: include: Obtaining historical status data of a heat source tower heat pump, various index values ​​of the historical status data under a preset applicability evaluation index system, and actual applicability evaluation results corresponding to the various index values; Using each index value of the historical state data under the applicability evaluation index system as input data of training data; Using the true applicability evaluation results corresponding to the index values ​​as output data of training data; Based on the input data and the output data, the neural network model is trained using a particle swarm algorithm to obtain a suitability evaluation model.

2. The method according to claim 1, characterized in that The method of training the neural network model based on the input data and the output data using a particle swarm algorithm to obtain a suitability evaluation model includes: Based on the input data, output data, randomly generated initial weights and initial thresholds, a neural network model is trained using a particle swarm algorithm to obtain optimal weights and optimal thresholds; According to the optimal weight and the optimal threshold, a suitability evaluation model is obtained.

3. The method according to claim 2, characterized in that The method of using the particle swarm algorithm to train the neural network model to obtain the optimal weight and the optimal threshold includes: Initialize the parameters of the particle swarm algorithm, including the number of particles in the particle swarm, the particle position and the number of iterations, as well as initialize the individual extreme value and the global optimal extreme value, and the initial position of each particle corresponds to a set of initial weights and initial thresholds of the neural network model; Inputting the input data of the training data into the neural network model to obtain prediction data corresponding to the input data; Calculating the fitness function value of each particle based on the square of the difference between the predicted data and the output data; Based on the fitness function value of each particle, the particle swarm algorithm is used to iteratively update the individual extreme value and the global optimal extreme value until the updated individual extreme value and the global optimal extreme value reach the convergence condition. At this time, the preset inertia weight update expression is used to update the initial weight and the initial threshold, and the updated initial weight is determined as the optimal weight, and the updated initial threshold is determined as the optimal threshold.

4. The method according to claim 3, characterized in that The adopting of a preset inertia weight update expression to update the initial weight and the initial threshold comprises: The inertia weight of the particle is updated using a preset inertia weight update expression to obtain an updated inertia weight; It is determined whether the maximum number of iterations has been reached. If so, the initial weight and the initial threshold are updated based on the updated inertia weight.

5. The method according to claim 4, characterized in that The updating of the initial weight and the initial threshold based on the updated inertia weight includes: Obtaining a learning rate of the neural network model according to the updated inertia weight; According to the learning rate of the neural network model, the initial weight and initial threshold are updated through a back propagation algorithm.

6. The method according to claim 3 or 4, characterized in that The inertia weight update expression is as follows: Among them, ω(t) represents the inertia weight at the tth iteration; ω max Represents the maximum value of inertia weight; ω min represents the minimum value of the inertia weight; t represents the current number of iterations; t max Represents the maximum number of iterations; e is a natural constant.

7. The method according to any one of claims 3 to 5, characterized in that: The update expression of the initial weight is as follows: w(t+1)=w(t)+η(P best (t)-w(t))+η(G best (t)-w(t)); The update expression of the initial threshold is as follows: Wherein, w(t+1) represents the initial weight of the neural network model at the t+1th iteration; w(t) represents the initial weight of the neural network model at the tth iteration; η represents the learning rate of the neural network model; P best (t) represents the individual extreme value at the tth iteration; G best (t) represents the global optimal extreme value at the t-th iteration; b(t+1) represents the initial threshold at the t+1-th iteration; b(t) represents the initial threshold at the t-th iteration; represents the output data at the t-th iteration; Represents the predicted data corresponding to the input data at the tth iteration.

8. The method according to claim 1, characterized in that The neural network model is a back propagation neural network model.

9. A system for constructing a suitability evaluation model, characterized in that: include: A data acquisition module, used to acquire historical status data of a heat source tower heat pump, various index values ​​of the historical status data under a preset applicability evaluation index system, and actual applicability evaluation results corresponding to the various index values; A training input module, used to use the index values ​​of the historical status data under the applicability evaluation index system as input data of training data; A training output module, used to use the true applicability evaluation results corresponding to the indicator values ​​as output data of training data; The model building module is used to train the neural network model based on the input data and the output data using a particle swarm algorithm to obtain a suitability evaluation model.

10. The system according to claim 9, characterized in that The model building module comprises: A model training submodule is used to train the neural network model using a particle swarm algorithm based on the input data, output data, randomly generated initial weights and initial thresholds to obtain optimal weights and optimal thresholds; The parameter adjustment submodule is used to obtain a suitability evaluation model according to the optimal weight and the optimal threshold.

11. The system according to claim 10, characterized in that The model training submodule includes: A parameter initialization unit, used to initialize the parameters of the particle swarm algorithm, the parameters including the number of particles in the particle swarm, the particle positions and the number of iterations, as well as initializing individual extreme values ​​and global optimal extreme values, and the initial position of each particle corresponds to a set of initial weights and initial thresholds of the neural network model; A model output unit, used to input the input data of the training data into the neural network model to obtain prediction data corresponding to the input data; A fitness calculation unit, used to calculate the fitness function value of each particle based on the square of the difference between the predicted data and the output data; A parameter updating unit is used to iteratively update the individual extreme value and the global optimal extreme value based on the fitness function value of each particle by using the particle swarm algorithm until the updated individual extreme value and the global optimal extreme value reach the convergence condition. At this time, a preset inertia weight update expression is used to update the initial weight and the initial threshold, and the updated initial weight is determined as the optimal weight, and the updated initial threshold is determined as the optimal threshold.

12. The system according to claim 11, characterized in that The parameter updating unit comprises: The weight updating subunit is used to update the inertia weight of the particle using a preset inertia weight updating expression to obtain an updated inertia weight; The parameter determination subunit is used to determine whether the maximum number of iterations has been reached. If so, the initial weight and initial threshold are updated based on the updated inertia weight.

13. The system of claim 12, wherein: The parameter determination subunit includes: A learning rate updating component, used to obtain a learning rate of the neural network model according to the updated inertia weight; The optimal parameter updating component is used to update the initial weight and initial threshold according to the learning rate of the neural network model through a back propagation algorithm.

14. The system according to claim 11 or 12, characterized in that The inertia weight update expression is as follows: Among them, ω(t) represents the inertia weight at the tth iteration; ω max Represents the maximum value of inertia weight; ω min represents the minimum value of the inertia weight; t represents the current number of iterations; t max Represents the maximum number of iterations; e is a natural constant.

15. The system according to any one of claims 11 to 13, characterized in that: The update expression of the initial weight is as follows: w(t+1)=w(t)+η(P best (t)-w(t))+η(G best (t)-w(t)); The update expression of the initial threshold is as follows: Wherein, w(t+1) represents the initial weight of the neural network model at the t+1th iteration; w(t) represents the initial weight of the neural network model at the tth iteration; η represents the learning rate of the neural network model; P best (t) represents the individual extreme value at the tth iteration; G best (t) represents the global optimal extreme value at the t-th iteration; b(t+1) represents the initial threshold at the t+1-th iteration; b(t) represents the initial threshold at the t-th iteration; represents the output data at the t-th iteration; Represents the predicted data corresponding to the input data at the tth iteration.

16. The system of claim 9, wherein: The neural network model is a back propagation neural network model.

17. A method for evaluating the suitability of a heat source tower heat pump, characterized in that: include: Get the status data of the heat pump in the heat source tower; According to the state data, using a suitability evaluation model to evaluate the suitability of the heat source tower heat pump, to obtain a suitability evaluation result of the heat source tower heat pump; Wherein, the applicability evaluation model is constructed using a heat source tower heat pump applicability evaluation model construction method as described in any one of claims 1-8.

18. The method according to claim 17, characterized in that The comprehensive evaluation index system of applicability includes: system performance index, energy-saving and environmental protection performance index and economic index; The system performance index includes one or more of the following: cooling capacity of the heat source tower heat pump, heating capacity of the heat source tower heat pump, winter heating performance coefficient of the heat source tower heat pump host, summer cooling performance coefficient of the heat source tower heat pump host and system comprehensive performance coefficient; The energy-saving and environmental protection index includes one or more of the following: carbon emission cost and system energy consumption coefficient; The economic indicators include one or more of the following: investment cost and operating cost; The status data includes one or more of the following: heating capacity, cooling capacity, air mass flow, water temperature on the unit outlet side, water temperature on the unit inlet side, host input power, compressor power consumption, tower pump side power consumption, tower side fan power consumption, fossil fuel consumption, basic carbon content of fossil fuels, carbon oxidation rate of fossil fuels, relative molecular mass ratio of carbon dioxide to carbon, power consumption of air-conditioning system, energy consumption, cooling demand of associated buildings, heating demand of associated buildings, equipment procurement cost, installation cost, land cost, construction cost, host cost, chiller cost, solution pump cost, operating solution cost and water cost.

19. A heat source tower heat pump suitability evaluation system, characterized in that: include: The operation data acquisition module is used to obtain the status data of the heat pump of the heat source tower; A suitability evaluation module, used to evaluate the suitability of the heat source tower heat pump according to the state data using a suitability evaluation model to obtain a suitability evaluation result of the heat source tower heat pump; Wherein, the applicability evaluation model is constructed using a heat source tower heat pump applicability evaluation model construction method as described in any one of claims 1-8.

20. The system of claim 19, wherein: The comprehensive evaluation index system of applicability includes: system performance index, energy-saving and environmental protection performance index and economic index; The system performance index includes one or more of the following: cooling capacity of the heat source tower heat pump, heating capacity of the heat source tower heat pump, winter heating performance coefficient of the heat source tower heat pump host, summer cooling performance coefficient of the heat source tower heat pump host and system comprehensive performance coefficient; The energy-saving and environmental protection index includes one or more of the following: carbon emission cost and system energy consumption coefficient; The economic indicators include one or more of the following: investment cost and operating cost; The status data includes one or more of the following: heating capacity, cooling capacity, air mass flow, water temperature on the unit outlet side, water temperature on the unit inlet side, host input power, compressor power consumption, tower pump side power consumption, tower side fan power consumption, fossil fuel consumption, basic carbon content of fossil fuels, carbon oxidation rate of fossil fuels, relative molecular mass ratio of carbon dioxide to carbon, power consumption of air-conditioning system, energy consumption, cooling demand of associated buildings, heating demand of associated buildings, equipment procurement cost, installation cost, land cost, construction cost, host cost, chiller cost, solution pump cost, operating solution cost and water cost.

21. An electronic device, characterized in that: include: at least one processor and memory; The memory and the processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, a method for constructing a heat source tower heat pump suitability evaluation model as described in any one of claims 1 to 8 or a method for evaluating the suitability of a heat source tower heat pump as described in any one of claims 17-18 is implemented.

22. A computing device readable storage medium, characterized in that: An execution program is stored thereon, and when the execution program is executed, a method for constructing a heat source tower heat pump suitability evaluation model as described in any one of claims 1 to 8 or a method for evaluating the suitability of a heat source tower heat pump as described in any one of claims 17-18 is implemented.

Images