JP5693257B2 - Heat source system configuration search apparatus, method and program thereof - Google Patents

Description

  The present invention relates to a heat source system configuration search apparatus, a method thereof, and a program for determining a combination of a heat source machine type capable of supplying a heat load given as a condition and a capacity ratio thereof in consideration of running cost.

  Conventionally, when designing a large-scale heat supply plant, it is necessary to supply the heat load given as a condition from among multiple types of heat source machines such as inverter turbo chillers, fixed speed turbo chillers, and air-cooled heat pumps. The planner determined the combination of heat source units considered to be optimal. Therefore, even under the same load conditions, the combination of heat source units adopted and the capacity ratio of each heat source unit differ depending on the plan implementer, and there are differences in economic efficiency and energy saving depending on the experience values of the plan implementer. It was.

  Patent Document 1 proposes a method for efficiently and rapidly drafting an operation plan in a large-scale heat supply plant when the operation plan is changed due to fluctuations in demand, equipment failure, or the like. ing. Specifically, a heat storage plan for a heat storage device is created by applying a genetic algorithm.

  In Patent Document 2, in the automatic generation method of fuzzy inference rules, parameter adjustment is divided into two stages, and in the first stage, a parameter value whose error is within a predetermined range is obtained using a genetic algorithm. A technique for further adjusting a parameter value using a steepest descent method or the like in the stage is disclosed.

JP-A-8-136174 JP 7-28649 A

However, the search for the optimal solution using the gene algorithm as disclosed in Patent Document 1 requires a large amount of calculation and has a drawback that it takes time to obtain the optimal solution.
The method disclosed in Patent Document 2 is a technique for obtaining a membership function used in fuzzy control using a genetic algorithm and a steepest descent method. This method is used as a technique for determining the configuration of a heat source system as it is. It cannot be diverted. Furthermore, since the steepest descent method requires a continuous space, it has a drawback that it is difficult to apply to calculations in which only discrete data exists.

  The present invention has been made in view of such circumstances, and efficiently combines a suitable combination and capacity ratio of heat source devices capable of supplying a heat load given as a condition from among a plurality of heat source device candidates. It is an object of the present invention to provide a heat source system configuration search apparatus, method thereof, and program that can be determined as follows.

In order to solve the above problems, the present invention employs the following means.
The present invention is a heat source system configuration search device for determining a combination of a heat source machine model capable of supplying a heat load given as a condition and its capacity ratio in consideration of running cost, and comprising a plurality of model heat source machines A plurality of combinations of heat source machine models capable of supplying the heat load and their capacity ratios among the candidates, a means for generating a gene population using each of the combinations as a gene, and a predetermined heat for each of the genes A means for calculating a running cost when supplying a load; a means for selecting a predetermined number in order from a gene having a low running cost; and setting the selected predetermined number of genes as an initial point in a search space, and a downhill simplex Means for obtaining one convergence point from the initial point using the method, and adopting the convergence point as an elite It performs cross and mutation and means for generating a new gene cluster, provides a heat source system configuration searching apparatus to search for a solution to a predetermined convergence condition is satisfied gene is searched.

According to the above configuration, a plurality of combinations of heat source machine models capable of supplying a heat load and their capacity ratios are set from among a plurality of heat source machine candidates, and a gene population is generated using each of these combinations as a gene. Is done. Subsequently, for each gene of the gene population, a running cost when a predetermined heat load is supplied is calculated, and the result of calculating the running cost is evaluated as the degree of fitness of the gene. Number selected. The selected predetermined number of genes are converted from the data in the gene format into the capacity ratio information of the heat source device, and one convergence point is searched by the downhill simplex method using the converted data. For example, based on the capacity ratio information (initial point) of the heat source device converted from the selected genetic data, calculation points that slightly change this capacity ratio information are set, and downhill simplex is calculated from these initial points and calculation points. One convergence point is searched by the method. This convergence point is converted from the capacity ratio information of the heat source machine into gene format data, and used as an excellent combination of heat source devices for creating the next generation gene. Specifically, crossover and mutation are performed based on this genetic data to generate a new gene population. The search for a solution is repeated until a gene that satisfies a predetermined convergence condition is found.
In this way, the genetic algorithm and the downhill simplex method are combined, and the area where the optimal solution is likely to exist is narrowed down using the genetic algorithm, and the solution in this area is narrowed down by the downhill simplex method. As a result, the number of searches can be greatly reduced as compared to the case of searching for a solution by a genetic algorithm alone. Further, even when a local solution is encountered, gene crossing and mutation are performed by the genetic algorithm, so it is possible to escape from the local solution and to search for the optimal solution reliably.

  The present invention is a heat source machine configuration search method for determining a combination of a heat source machine model capable of supplying a heat load given as a condition and a capacity ratio thereof in consideration of running cost, and the computer is configured from a plurality of machine models. A plurality of combinations of heat source machine models and capacity ratios of the heat source machines that can supply the heat load from among the heat source machine candidates, and generating a gene population using each of the combinations as genes, and for each of the genes, A step of calculating a running cost when supplying a predetermined heat load, a step of selecting a predetermined number in order from genes having a low running cost, and setting the selected predetermined number of genes as an initial point in a search space. The process of obtaining one convergence point from the initial point using the downhill simplex method, and adopting the convergence point as an elite Rutotomoni by performing cross and mutation execute a step of generating a new gene cluster, a predetermined convergence condition is satisfied gene provides a heat source system configuration search method of searching for solutions to be searched.

  The present invention is a heat source machine configuration search program for determining a combination of a heat source machine model capable of supplying a heat load given as a condition and a capacity ratio thereof in consideration of running cost, and includes a plurality of models. A plurality of combinations of heat source machine models capable of supplying the heat load and capacity ratios among the heat source machine candidates, a process of generating a gene population using each of the combinations as data in a gene format, and each of the genes In the search space, a process for calculating a running cost when supplying a predetermined heat load, a process for selecting a predetermined number in order from the gene with the lowest running cost, and the selected predetermined number of genes as an initial point in the search space Processing to obtain a corresponding convergence point from the initial point using the downhill simplex method, and to elite the convergence point In addition, the computer system executes a process of generating a new gene population by crossing and mutating, and searching for a heat source system configuration that searches for a solution until a gene that satisfies a predetermined convergence condition is searched. Provide a program.

  According to the present invention, it is possible to efficiently determine a suitable combination and capacity ratio of heat source devices capable of supplying a heat load given as a condition from among a plurality of heat source device candidates.

It is the block diagram which showed the hardware constitutions of the heat-source system structure search apparatus which concerns on one Embodiment of this invention. It is the flowchart which showed the process sequence performed by the heat-source system structure search apparatus which concerns on one Embodiment of this invention. It is the figure which showed the example of the combination of the heat source machine set as a gene, and capacity | capacitance ratio. It is the figure which showed an example of the genetic data structure based on the combination shown in FIG. It is explanatory drawing for demonstrating briefly the procedure of the downhill simplex method. It is a figure for demonstrating gene crossing. It is a figure for demonstrating a mutation. It is the flowchart which showed the process sequence of the driving | running pattern search method. It is the figure which showed the conceptual diagram of the driving | operation combination of a heat source machine when the cooling load of 200 kW is selected. It is the figure which showed an example of the relationship between a driving | operation combination and an electric power demand. It is the figure which showed an example of the performance characteristic of the air cooling chiller. It is the figure which showed an example of the performance characteristic of an absorption refrigerator. It is the figure which showed an example of the performance characteristic of the inverter turbo refrigerator. It is the figure which showed an example of the performance characteristic of a fixed speed turbo refrigerator. It is the figure which showed an example of the electric power charge. It is the figure which showed an example of the relationship between a driving | running combination and running cost. It is a figure for demonstrating the effect of the heat-source system structure search apparatus which concerns on one Embodiment of this invention. It is the figure which showed an example of the annual pattern of a cooling load.

  Hereinafter, a heat source system configuration searching device according to an embodiment of the present invention will be described with reference to the drawings. The heat source system configuration search apparatus according to the present embodiment is an apparatus that searches for an optimum solution for a combination of heat source units that minimize the running cost and its capacity ratio with respect to a thermal load given as a condition. In the present embodiment, one of the main features is that the optimum solution is searched by a method combining a genetic algorithm and a downhill simplex method.

In general, the genetic algorithm has the advantage that even if it falls into the local solution, it can escape from the local solution by mutation etc., but the search for the optimal solution is based on the probabilistic calculation of gene crossover and mutation, There is a drawback that the number of calculations becomes enormous.
The downhill simplex method searches the optimal solution while following the shape of the search space, so if the calculation start point is good, it has the advantage that the optimal solution can be reached with a small number of calculations, but when it falls into a local solution It has the disadvantage that it is difficult to escape.
In the present embodiment, the optimal solution is searched using a novel algorithm that reliably searches for an optimal solution while combining the genetic algorithm and the downhill simplex method as described above, and suppressing the number of calculations.

Hereinafter, a specific apparatus configuration and processing procedure will be described.
FIG. 1 is a diagram illustrating a hardware configuration of a heat source system configuration searching apparatus 10 according to the present embodiment. As shown in FIG. 1, a heat source system configuration search device 10 is a computer system, and includes a CPU (Central Processing Unit) 11, a main storage device 12 such as a RAM (Random Access Memory), an auxiliary storage device 13, a keyboard and a mouse. And the like, an output device 15 such as a display and a printer, and a communication device 16 that exchanges information by communicating with external devices.

  The auxiliary storage device 13 is a computer-readable recording medium, such as a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, or a semiconductor memory. The auxiliary storage device 13 stores various programs (for example, a heat source system configuration search program, an operation pattern search program, etc.), and the CPU 11 reads the program from the auxiliary storage device 13 to the main storage device 12 and executes it. The optimal solution is searched by the above.

FIG. 2 is a flowchart showing a processing procedure executed by the heat source system configuration searching device 10. Hereinafter, each process will be described with reference to FIG.
In this case, inverter turbo chillers, fixed speed turbo chillers, air cooling heat pumps, gas absorption chillers, steam absorption chillers, heat recovery heat pumps, EHP (electric heat pump packages), and GHP (gas heat pumps) are listed as heat source candidates. ing. In the following heat source system configuration search method, the combination of the heat source units that are suitable for supplying the heat load given as a condition and the capacity ratio of each model are used as variables, and the heat source unit that has the minimum running cost. The combination and the capacity ratio Φ (i) of each model are searched for as an optimal solution. In other words, this is expressed as follows:

Capa (i) = Max_Fuka * Φ (i) * Factor (1)
Equip_Cap (i) = Capa (i) / N (i) (2)

  Where Capa (i) is the total capacity (kW) of each heat source machine, Equip_Cap (i) is the equipment capacity (kW), N (i) is the number of heat sources, Φ (i) is the capacity ratio of the heat source machine, and ΣΦ (i) = 1, Max_Fuka is the maximum load processed by the heat source unit, and Factor is the equipment capacity margin, for example, about 1.0-1.2. The N (i) is restricted so that the equipment capacity Equip_Capa (i) is equal to or greater than the allowable minimum capacity (for example, 200 rt or more for a turbo chiller and 100 RT or more for an absorption chiller).

First, in step SA1 in FIG. 2, a plurality of genes (individuals) are set, and a gene group is set. Specifically, from among the above-mentioned multiple types of heat source devices listed as heat source device candidates, the types of heat source devices that can supply the heat load given as a condition (here, 1000 RT) and their capacity ratio A plurality of combinations are arbitrarily determined, and a gene population is generated with each of the combinations as a gene.
FIG. 3 shows an example of a combination of heat sources set as genes and a capacity ratio, and FIG. 4 shows an example of a gene data structure based on the combination shown in FIG.
In the genes shown in FIGS. 3 and 4, the capacity ratio Φ (i) of the inverter turbo chiller with respect to the given heat load 1000RT is 30%, the capacity ratio Φ (i) of the air-cooled heat pump is 50%, and the gas absorption refrigeration A combination in which the capacity ratio Φ (i) of the machine is 20% is set. In the gene data structure shown in FIG. 4, the set capacity ratio% of each heat source machine is expressed by 7 bits per model.

  Next, in step SA2, a running cost is obtained for each gene set in step SA1 using the driving pattern search method. Details of the driving pattern search method will be described later. When the running cost is calculated for each gene, it is determined in step SA3 whether there is a gene that satisfies the convergence condition. As a result, when there is a gene that satisfies the convergence condition, the processing is terminated with the gene as an optimal solution. On the other hand, if there is no gene that satisfies the convergence condition, the process proceeds to step SA4, the running cost calculated in step SA2 is evaluated as the fitness of the gene, and a predetermined number (for example, N) in the order of good evaluation results. Are selected as elite. In other words, a predetermined number is selected in order from the gene with the lowest running cost, and the selected predetermined number of genes is set as the elite.

  Next, in step SA5, N genes selected as elite are set as initial points in the search space, and one convergence point point is calculated from the initial points using the downhill simplex method. The downhill simplex method is a method in which a polyhedron is formed by using only information of an evaluation function value in a search space and progresses toward an optimum value in the search space. Specifically, a polyhedron having N initial points set in the search space as end points is configured. Then, the convergence point is calculated by sequentially updating the worst point in the polyhedron by operations such as reflection, enlargement, reduction, and contraction as shown in FIG.

More specifically, one piece of genetic information selected as an elite is converted into device configuration ratio data. Next, a plurality of calculation points are set by minutely changing the device configuration data of the elite gene, and a running cost is calculated for each calculation point. Subsequently, a search space for the downhill simplex method is formed around a point corresponding to the genetic information selected by the elite and a plurality of set calculation points, and a direction in which the running cost decreases is searched. This search is performed by the following operation.
First, a point having the worst running cost among a plurality of calculation points set in the search space is specified as the worst point Xh, and the worst point Xh is replaced with the following Xr. (Operation 1 in FIG. 5).

  Xr = (1 + α) Xg−αXh

Here, α is a reflection coefficient of α> 0, and Xg is the center of gravity of all other points Xj except the worst point Xh.
Next, Xr is replaced with the following Xe along the Xg-Xr direction (operation 2 in FIG. 5).

  Xe = γXg + (1-γ) Xr

Here, γ is an expansion coefficient of γ> 0.
Next, Xe is replaced with the following Xc (operation 3 in FIG. 5).

  Xc = γXg + (1-γ) Xe

  Then, the entire simplex is reduced in half in the direction of Xl.

Xi = 0.5 (Xl + Xi)
i = 1, 2,... N

  Here, Xl is the best point, that is, the point of the gene with the lowest running cost.

  With respect to the points thus obtained, the running cost is again calculated using the driving pattern search method, the worst point is identified again, and the above operations 1 to 4 are sequentially processed. Then, by repeating this operation, the convergence point is finally calculated.

When the convergence point is obtained by the downhill simplex method, in step SA6, the convergence point information is converted into the gene data structure shown in FIG. 4 and transferred to the gene algorithm. In the genetic algorithm, the gene at the convergence point obtained by the downhill simplex method is treated as an elite gene, and crossing and mutation are performed.
The gene crossing is performed, for example, by exchanging a part of the gene expressed by the bit length as shown in FIG. As shown in FIG. 7, the mutation determines a bit (part) to be mutated by a random number and changes the value of the bit. When a new gene population is generated by performing such an operation, the processing after step SA2 is performed on this gene population. Steps SA2 to SA7 are repeated until the convergence condition is satisfied and an optimal solution is obtained.

Next, a running cost calculation method using the driving pattern search method performed in step SA2 will be described with reference to FIG. FIG. 8 is a flowchart showing a processing procedure of the driving pattern search method.
In the operation pattern search method, the minimum running cost when the operation is performed using the heat source system including the combination of heat source units determined by each gene is searched. For example, taking the gene shown in FIG. 3 as an example, a heat source system including one inverter turbo refrigerator (300RT), 20 air-cooled heat pumps (25RT), and two gas absorption refrigerators (100RT) is set. Therefore, an operation pattern that minimizes the annual running cost in this heat source system is searched for as an optimal solution.

  First, the annual pattern of the cooling load and the annual pattern of the heating load are read from the auxiliary storage device 13 (see FIG. 1) (step SB1 in FIG. 8). This annual pattern can be obtained, for example, by accumulating load data in the past and statistically processing the accumulated load data. Here, the annual pattern of the cooling load and the annual pattern of the heating load may be, for example, as shown in FIG. 18, in which a load for 24 hours in a typical day of the month is shown for each month. Alternatively, a representative daily load of the month (August in the example of FIG. 18) in which the highest load is recorded may be shown. In the present embodiment, a case where an annual pattern showing a typical daily load of each month shown in FIG. 18 is applied will be described as an example.

  Next, the cooling load for the first hour (for example, 0: 00-0: 59 in January) is selected from the annual pattern of the cooling load (step SB2 in FIG. 8), and this cooling source system uses this cooling load. All the operation combinations of the heat source machines that can supply the load are extracted (step SB3 in FIG. 8). In FIG. 9, the conceptual diagram of the driving | operation combination of a heat source machine when the cooling load of 200 kW is selected is shown. In FIG. 9, for example, the heat source unit 1 is an inverter turbo refrigerator, the heat source unit 2 is an air-cooled heat pump, and the heat source unit 3 is a gas absorption refrigerator.

  Next, from among the operating combinations of the heat source units, the combinations in which the cooling load (200 kW) falls within the range of 20% or more and 100% or less of the maximum output of the operating combination remain, and combinations outside this range are excluded (FIG. 8). Step SB4). Thereby, for example, the operation combinations “3”, “5”, and “7” in FIG. 9 are excluded.

  Next, a power demand is calculated for each of the operation combinations of the heat source machines remaining in step SB3, and operation combinations in which this power demand exceeds the target value are excluded (step SB5 in FIG. 8). FIG. 10 shows an example of the relationship between the driving combination and the power demand. According to FIG. 10, the driving combinations “1”, “8”, and “X” are excluded.

Next, the running cost per hour based on the weight charge of power and gas is calculated for each of the remaining operation combinations (step SB6 in FIG. 8).
Here, in calculating the running cost, various data necessary for calculating the running cost are stored in the main storage device 13 (see FIG. 1) in advance. Examples of stored data include (A) partial load characteristics of heat source machine candidates, (B) calculation formula for cooling water temperature, (C) power charge system for power cost, (D) calculation formula for pump power, (E) The setting method of the auxiliary machine system of a heat source system etc. are mentioned. Each will be described in detail below.

(A) Partial load characteristics of heat source machine candidates As shown in FIGS. 11 to 14, the performance characteristics of each heat source machine are different. By preparing the performance characteristics of each heat source machine that is a candidate heat source machine, it is possible to assign each heat source machine to a load zone where each heat source machine can operate with high efficiency, and create an optimal operation pattern be able to.

(B) Cooling water temperature calculation formula The cooling water temperature is calculated based on the design conditions by calculating the NTU of the operation point of the cooling tower using the refrigerator load, the outside air wet bulb temperature, the air volume ratio to the rated conditions, and the water volume ratio as parameters. By obtaining the outlet temperature that is the same as that of the NTU by repeated temperature calculation, the performance evaluation for the outdoor air wet bulb temperature of the heat source is calculated. For example, the estimated value of the outlet cooling water temperature t2 with respect to the inlet cooling water temperature t1 of the cooling tower is obtained by calculating the outlet temperature at which the design condition NTU and the operation condition NTU1 match by numerical integration as shown in the following equation (3). It can be calculated by repeated calculation. In evaluating the running cost, it is possible to improve the accuracy of the running cost by further considering the amount of water used in the cooling tower.

  In equation (3), NTU is the basic unit of movement, Cpw is the specific enthalpy of water (kJ / kg. ° C), hs is the inlet air enthalpy (kJ / kg), h is the outlet air enthalpy (kJ / kg), and tw1 is Cooling tower inlet cooling water temperature (° C.), tw 2 is the cooling tower outlet cooling water temperature (° C.), L is the cooling water amount (kg / h), Ka is the enthalpy reference energy transfer coefficient, and V is the air flow rate.

(C) Electricity cost system for electricity costs In order to evaluate the electricity cost in accordance with the billing fee system of the electric power company, daytime electricity, nighttime electricity, heavy load time, heat storage fee, basic fee (demand), self-supplied electricity Etc., and calculate the power cost according to the billing system. Thereby, the operation cost of the heat source facility can be obtained with high accuracy. An example of the power charge is shown in FIG.

(D) Calculation formula for pump power For heat source equipment with a pump inverter (for example, inverter refrigerator), the pump discharge pressure is calculated from the flow rate load factor as shown in the following formulas (4) and (5). And calculate the power.

Here, P is the pump head (kPa), a, b and c are constants depending on the piping system, Q is the flow rate (m ³ / h), PW is the pump power (kW), and ρ is the water specific gravity (kg / m ³ ) , G is the acceleration of gravity, and η is the pump efficiency.

(E) Setting method of auxiliary system of heat source system Equipment performance parameters (power consumption and gas of refrigerators and boilers, etc.) are automatically calculated by making database of manufacturer's equipment characteristics. For example, the chilled water pump calculates up to the primary pump of a two-pump system, the chilled water pressure loss is set to 0.2 MPa, and the cooling water pump automatically sets the pump power with the chilled water pressure loss of 0.25 MPa.

  Then, using the information as described above, the hourly running cost is calculated using the following equation (6).

In the above equation (6), Cost (m, h) is a running cost (yen / h), which is a metered rate base power cost + gas cost + water supply and drainage cost. PEq (k, m, j) is the power consumption (kWh) of heat source k, Puser (m, h) is the consumer power load (kWh), GEq (k, m, j) is the gas consumption of heat source k Amount (Nm ³ / h), Gcgs (l, m, h) is the water consumption of the cogeneration unit 1 (m ³ ), WEq (k, m, j) is the water consumption of the heat source k (m ³ ), Wcgs (l, m, h) is the water consumption (m ³ ) of the cogeneration unit 1, Punit is the unit price for electricity (yen / kWh), Gunit is the unit price for gas (yen / m ³ ), Wunit is the unit price for water and sewage ( Yen / m ³ ), n is the number of heat source units, q is the number of cogeneration units, and m and h are months and hours.

  In the above equation (6), the equipment power consumption PEq (k, m, h), the equipment gas consumption GEq (k, m, h), and the equipment water consumption WEq (k, mh) It is a function of the load η (%) of the heat source unit, the outside air temperature Tg, or the cooling water temperature Tw, and is expressed by a function φ (η, Tw) that formulates the performance characteristic curve of the device.

  In this way, when the running cost per hour is calculated for all the remaining combinations in step SB6 in FIG. 8, the driving combination showing the minimum running cost is selected as the optimum operating time zone. It selects as a driving | operation pattern, this driving | operation combination and running cost are matched and preserve | saved (step SB7 of FIG. 8). FIG. 16 shows an example in which the driving combination “6” is selected as the driving combination indicating the minimum running cost.

  Next, it is determined whether or not the calculation of the minimum running cost has been completed for the annual pattern of the cooling load (step SB8 in FIG. 8). If the calculation has not been completed (NO in step SB8 in FIG. 8). )), The cooling load of the value of the refrigeration load at the next time (for example, 1: 00-1: 59 in January) is set (step SB9 in FIG. 8), and the cooling load after step SB3 is set for this cooling load. Process. Then, by repeatedly performing the processes of steps SB3 to SB9, when it is determined in step SB8 of FIG. 8 that the calculation of the minimum running cost is completed for the annual pattern of the cooling load, the process of the running cost related to the cooling load is terminated. Then, according to the same procedure, the annual minimum running cost is calculated for the heating load (step SB10 in FIG. 8), and the process ends.

  The minimum running cost calculation process described above is performed for each heat source system having a device configuration specified by each gene set in step SA1 (SA7) in FIG. 2, so that the minimum running cost for each gene can be reduced. Will be calculated.

  As described above, according to the heat source machine configuration system and method therefor according to the present embodiment, the area where the optimal solution is likely to exist is narrowed down using the genetic algorithm, and further, the downhill simplex method is used to Since the optimum solution is narrowed down using the genes in the area searched by the algorithm, it is possible to obtain an effect that the optimum solution can be reliably obtained with a smaller number of searches compared to the gene algorithm.

FIG. 17 shows the results of searching for the optimal solution using the heat source system configuration search device according to the present embodiment, the results of searching for the optimal solution with the genetic algorithm alone, and the optimal solution with the downhill simplex method alone. A comparison is made with the results of the search. Here, as a result, the combination of the heat source unit, the capacity ratio, and the running cost in which the running cost is minimized when the annual pattern of the cooling load as shown in FIG. 18 is applied are shown.
As shown in FIG. 17, it was verified that the running cost by the heat source machine configuration searched by the heat source system configuration search apparatus according to the present embodiment is the lowest. In addition, the number of searches was 280, which was not as high as the 187 downhill simplex method, but it was verified that the number of searches was significantly reduced compared to the 1320 genetic algorithm.

10 Heat Source System Configuration Search Device 11 CPU
12 Main storage device 13 Auxiliary storage device 14 Input device 15 Output device 16 Communication device

Claims (3)

A heat source system configuration search device that determines a combination of a heat source machine model capable of supplying a heat load given as a condition and a capacity ratio thereof in consideration of running cost,
A plurality of combinations of heat source machine models capable of supplying the heat load and capacity ratios among heat source machine candidates consisting of a plurality of models, and means for generating a gene population using each of the combinations as a gene;
For each of the genes, means for calculating a running cost when supplying a predetermined heat load;
Means for selecting a predetermined number in order from the gene having a low running cost;
Means for setting the selected predetermined number of genes as an initial point in a search space, and obtaining one convergence point from the initial point using a downhill simplex method;
Adopting the convergence point as an elite, and means for generating a new gene population by performing crossover and mutation,
A heat source system configuration search device that searches for a solution until a gene that satisfies a predetermined convergence condition is searched. A heat source device configuration search method for determining a combination of a model of a heat source device capable of supplying a heat load given as a condition and a capacity ratio thereof in consideration of a running cost,
Computer
Setting a plurality of combinations of heat source machine models capable of supplying the heat load and capacity ratios among heat source machine candidates consisting of a plurality of models, and generating a gene population using each of the combinations as genes; and
For each of the genes, calculating a running cost when supplying a predetermined heat load;
Selecting a predetermined number in order from the gene having a low running cost;
Setting the selected predetermined number of selected genes in the search space as an initial point, and determining one convergence point from the initial point using a downhill simplex method;
Adopting the convergence point as an elite and performing crossing and mutation to generate a new gene population,
A heat source system configuration search method that searches for a solution until a gene that satisfies a predetermined convergence condition is searched. A heat source machine configuration search program for determining a combination of a heat source machine model capable of supplying a heat load given as a condition and its capacity ratio in consideration of running cost,
A plurality of combinations of heat source machine models capable of supplying the heat load and capacity ratios among the heat source machine candidates consisting of a plurality of models, and a process of generating a gene population using each of the combinations as a gene;
For each of the genes, a process of calculating a running cost when supplying a predetermined heat load;
A process of selecting a predetermined number in order from the gene having a low running cost;
A process for setting the selected predetermined number of selected genes as an initial point in a search space and obtaining one convergence point from the initial point using a downhill simplex method;
Adopting the convergence point as an elite, causing the computer to execute a process of generating a new gene population by performing crossover and mutation,
A heat source system configuration search program that searches for a solution until a gene that satisfies a predetermined convergence condition is searched.

Images