Planning device
The planning device optimizes the hydrogen supply chain using distributed computing and quantum resources to address the computational challenges of decarbonization, ensuring efficient resource allocation and planning.
Patent Information
- Application Number
- JP2023075156
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-04-28
- Publication Date
- 2024-11-08
AI Technical Summary
Formulating a plan for the entire hydrogen supply chain to promote decarbonization poses a significant challenge due to the enormous amount of calculation required for prediction, optimization, and learning.
A planning device that optimizes the entire resource supply chain by connecting to power procurement, fuel material production, storage, and delivery devices, utilizing multiple computer resources with different characteristics to distribute and optimize manufacturing, storage, supply, and delivery plans, and integrating quantum computing for enhanced calculation capabilities.
Efficiently optimizes the hydrogen supply chain planning, reducing computational burden and ensuring optimal resource allocation across various stages, thereby facilitating decarbonization efforts.
Smart Images

Figure 2024159280000001_ABST
Abstract
Description
[Technical field]
[0001] An embodiment of the present invention relates to a planning device that performs planning for an entire resource supply chain. [Background technology]
[0002] Hydrogen energy is gaining attention as a new clean energy source. Using electricity and water as inputs, hydrogen is produced as a fuel substance by a hydrogen production device. Using hydrogen as input, electricity and heat are produced by a hydrogen power generation device. In other words, using a hydrogen production device, hydrogen storage device, and hydrogen power generation device (hereafter referred to as a hydrogen system), electricity can be converted into hydrogen and then converted back into electricity. By connecting this group of devices to a power grid, it becomes possible to absorb electricity from the power grid when there is an excess supply from other power generation facilities (in this case, this becomes electricity demand), and to provide electricity to the power grid when there is an excess demand (in this case, this becomes electricity supply), thereby contributing to the stabilization of the power grid.
[0003] Attention is also being drawn to renewable energy power generation facilities derived from natural energy (hereinafter referred to as renewable energy power generation facilities). There is a movement in countries around the world to significantly increase the proportion of renewable energy power generation facilities in their domestic power sources, and expectations are growing for hydrogen systems as a means of stabilizing renewable energy power generation, which fluctuates greatly depending on weather conditions.
[0004] On the other hand, decarbonization through the utilization of hydrogen requires a consistent approach that covers not only the time of use but also the time of production, storage, transportation, etc. Therefore, it would be desirable to formulate a plan that covers the entire hydrogen supply chain, but the scale of calculations required for prediction, optimization, learning, etc. would be enormous. [Prior art documents] [Patent documents]
[0005] [Patent Document 1] International Publication No. 2020 / 196889 [Patent Document 2] Special Publication No. 2015-517144 Summary of the Invention [Problem to be solved by the invention]
[0006] As described above, in order to promote decarbonization by utilizing hydrogen as a resource, there is a problem that a huge amount of calculation is required to formulate a plan for the entire resource supply chain. The planning device of the embodiment has been made to solve this problem, and provides a planning device that can optimize the plan for the entire resource supply chain. [Means for solving the problem]
[0007] The planning device of the embodiment is connected to a power procurement source, a fuel material manufacturing apparatus, a fuel material storage apparatus, a fuel material supply apparatus, and a fuel material delivery apparatus, and is a planning device that optimizes an overall plan including a production plan in which the fuel material manufacturing apparatus receives power supply from the power procurement source and produces fuel material, a storage plan in which the fuel material storage apparatus stores fuel material, a supply plan in which the fuel material supply apparatus supplies fuel material, and a delivery plan in which the fuel material delivery apparatus delivers fuel material. The system comprises an information receiving unit that receives power information including the amount of power supplied from a power procurement source, production information including the production amount in a fuel material production device, storage information including the storage amount in a fuel material storage device, supply information including the supply amount from a fuel material supply device, and delivery information including the delivery amount from a fuel material delivery device; a setting unit that sets calculation conditions for optimizing the overall plan based on the power information, production information, storage information, supply information, and delivery information; a problem division unit that divides problems related to the production plan, storage plan, supply plan, and delivery plan that make up the overall plan into two or more problem groups based on specified conditions; a resource allocation unit that assigns two or more computer resources having different characteristics to each of the divided problem groups in accordance with the characteristics of the problem groups; a problem transmission / reception unit that transmits the problem groups to which the computer resources have been assigned and the calculation conditions to the computer resources and receives the calculation results from the computer resources; and a combination unit that combines the calculation results received for each problem group to generate an overall plan. [Brief description of the drawings]
[0008] [Figure 1] 1 is a block diagram showing a configuration example of an energy system for which a planning device of a first embodiment creates a plan. [Diagram 2] 1 is a block diagram showing a configuration of a planning device according to a first embodiment. [Diagram 3] FIG. 2 is a block diagram showing variables in planning by the planning device of the first embodiment. [Figure 4] 4 is a flowchart showing the overall operation of the planning device of the first embodiment. [Diagram 5] 5 is a flowchart showing a division operation of the planning device of the first embodiment. [Figure 6A] FIG. 4 is a conceptual diagram showing an example of a division operation of the planning device of the first embodiment. [Figure 6B] FIG. 4 is a conceptual diagram showing an example of a division operation of the planning device of the first embodiment. [Figure 7] FIG. 4 is a conceptual diagram showing an example of a division operation of the planning device of the first embodiment. [Figure 8] FIG. 4 is a conceptual diagram showing an example of an allocation operation of the planning device of the first embodiment. [Figure 9] 13 is a flowchart showing an operation example of the planning device of the second embodiment. [Figure 10] 13 is a flowchart showing an operation example of the planning device of the third embodiment. [Figure 11] 13 is a flowchart showing an operation example of the planning device according to the fourth embodiment. [Figure 12] 13 is a flowchart showing an operation example of the planning device according to the fifth embodiment. [Figure 13] 13 is a flowchart showing an operation example of the planning device according to the sixth embodiment. [Figure 14] FIG. 23 is a conceptual diagram showing an example of the operation of the planning device according to the seventh embodiment. [Figure 15] FIG. 13 is a block diagram showing a configuration of a planning device according to an eighth embodiment. [Figure 16] 13 is a flowchart showing an operation example of the planning device according to the eighth embodiment. [Figure 17]FIG. 23 is a diagram illustrating an example of screen display information of the planning device according to the ninth embodiment. [Figure 18] FIG. 23 is a diagram illustrating an example of screen display information of the planning device according to the ninth embodiment. [Figure 19] FIG. 23 is a diagram illustrating an example of screen display information of the planning device according to the ninth embodiment. [Figure 20] FIG. 23 is a diagram illustrating an example of screen display information of the planning device according to the ninth embodiment. [Figure 21] FIG. 23 is a diagram illustrating an example of screen display information of the planning device according to the ninth embodiment. [Figure 22] FIG. 23 is a diagram illustrating an example of screen display information of the planning device according to the ninth embodiment. [Diagram 23] FIG. 23 is a diagram illustrating an example of screen display information of the planning device according to the ninth embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0009] (First embodiment) To achieve decarbonization through the use of hydrogen as a fuel material, a consistent approach is required, including not only the time of use but also the time of production, storage, and transportation. It would be preferable to formulate a plan that covers the entire supply chain of resources such as hydrogen, but the scale of the problem to be calculated would be enormous. Therefore, the planning device of the embodiment realizes distributed processing that is optimal for calculations in planning.
[0010] (Energy system configuration) As shown in Fig. 1, the energy system 1 of the embodiment includes a hydrogen production system 50 including a hydrogen production device 20, a storage device 22, a compression device 24, a supply device 26, a renewable energy power generation device (renewable energy power generation device) 44, and a storage complex device 46, and a planning device 10 interconnected with these. The energy system 1 may have a transport device 28 that transports hydrogen supplied from the supply device 26 and a demand device 30 that receives a supply of hydrogen from the transport device 28 under its control, and may be further interconnected with the planning device 10. The planning device 10 may further be connected to a wholesale electricity trading market 52, a relative power 54, and a forecasting service 60. In this embodiment, an example will be described in which the fuel material (resource) to be managed is hydrogen, but the present invention is not limited to this.
[0011] The planning device 10 is also connected to one or more computer resources 70. Planning for the entire supply chain of resources such as hydrogen involves a wide variety of problems to be calculated, resulting in a huge number of problems. For this reason, it is not realistic to handle all of the problems with a single computer resource, and it is preferable to distribute the optimization calculations for the problems to multiple computer resources. The planning device 10 of the embodiment divides the problems to be calculated and causes the multiple computer resources 70 to perform the calculations.
[0012] The computer resource 70 can be realized by a conventional computer such as a von Neumann type computer, an FPGA (Field-Programmable Gate Array: a circuit array that can be reprogrammed on the spot), an ASIC (Application Specific Integrated Circuit: an integrated circuit for a specific application), a GPU (Graphics Processing Unit: a computing device specialized for image processing), or a combination thereof, or a quantum computer (a gate type including light quantum, an annealing type, etc.). Here, in the conventional computer, a method that takes into account quantum phenomena (simulated bifurcation algorithm) or the like may be adopted. The computer resource 70 that distributes the calculation is not limited to a computer as a hardware resource, and may include a calculation tool as a software resource. The computer resource 70 may be located at a location physically different from the planning device 10, or may be integrated with the planning device 10 as a computer system.
[0013] The computer resource 70 has a function of performing optimization calculations on the problem transmitted by the planning device 10 and returning the calculation results. The computer resource 70 may have a function of encrypting the calculation results to be returned to the planning device 10.
[0014] The planning device 10 creates a hydrogen supply plan that optimizes the amount of power supplied to the hydrogen production device 20, the source and amount of power supplied, etc., based on parameters acquired from the hydrogen production device 20, the storage device 22, the compression device 24, the supply device 26, the power grid 40, the renewable energy power generation device 44, and the storage battery device 46. The planning device 10 may further acquire parameters from the transportation device 28, the demand device 30, the wholesale power trading market 52, and the relative power 54 to create the hydrogen supply plan, etc. The planning device 10 formulates the plan in cooperation with computer resources 70.
[0015] The hydrogen production device 20 (production device) is an element that uses electric power to produce hydrogen as a fuel substance. The hydrogen production device 20 produces hydrogen using electric power supplied from one or more of the power grid 40, the renewable energy power generation device 44, and the storage battery device 46. The hydrogen production device 20 provides the planning device 10 with information such as the average electric power received per unit time and the amount of hydrogen produced per unit time. The hydrogen production device 20 produces hydrogen based on the plan created by the planning device 10.
[0016] The fuel substance is not limited to hydrogen. For example, it may be ammonia, methane, methanol, naphtha, gasoline, kerosene, jet fuel (SAF), diesel, heavy oil, ethanol, ethylene, LPG, carbon monoxide, etc. These fuel substances may be converted from hydrogen after hydrogen production, storage, or transportation, or may be fuel substances produced using a device that produces these fuel substances directly from electricity instead of a hydrogen production device. In addition, it may also be mixed with imported fuel substances or fuel substances produced by other means.
[0017] The storage device 22 is an element that stores hydrogen. The storage device 22 stores the hydrogen produced by the hydrogen production device 20. The storage device 22 provides the planning device 10 with information such as the amount of stored hydrogen and the amount of hydrogen supplied to the compression device 24.
[0018] The compressor 24 is an element that uses electric power to compress the hydrogen stored in the storage device 22. The compressor 24 can compress hydrogen to convert gaseous hydrogen into compressed hydrogen.
[0019] The supply device 26 is an element that supplies (fills) the hydrogen compressed by the compression device 24 to the transport device 28. The supply device 26 is exemplified by a hydrogen station that supplies hydrogen to a demand device 30, which is a hydrogen user, via the transport device 28. The supply of fuel material may also include supply to a buffer tank, supply to a trailer or a cardle, supply to a fuel consumer (e.g., a fuel supply station), a fuel cell vehicle, a fuel cell, a mixed-fuel turbine (e.g., a hydrogen mixed-fuel turbine), or a mono-fuel turbine (e.g., a hydrogen mono-fuel turbine).
[0020] The transportation device 28 is an element that transports hydrogen supplied from the supply device 26 to the demand device 30. The transportation device 28 is, for example, a transportation vehicle such as a tanker truck, a cargo ship, a railway, a pipeline, etc. The transportation device 28 provides the planning device 10 with information such as the amount of hydrogen stored and the amount of hydrogen transported in the transportation device 28.
[0021] The demand device 30 is an element that uses hydrogen. The demand device 30 includes, for example, a storage device that stores hydrogen for a certain period of time, a supply device that supplies hydrogen received from the transport device 28 to other elements, a power generation device that generates electricity using the received hydrogen, and a hot water / steam production device that produces hot water or steam using the received hydrogen. An example of a supply device as the demand device 30 is a hydrogen station that fills FCVs, FC forklifts, etc. An example of a power generation device as the demand device 30 is an FC or a hydrogen gas turbine, etc. An example of a hot water / steam production device as the demand device 30 is a hydrogen boiler, etc.
[0022] The power system 40 is an element that procures power from a wholesale power trading market 52 such as a wholesale power market or a bilateral power contract partner, or a bilateral power 54. The power system 40 supplies power to each element of the energy system 1.
[0023] The power system 40 may be a power system having a plurality of devices operated by a power company. The power system 40 generally sells power, but may also be in a position to purchase power from other power companies. The power system 40 provides the planning device 10 with information such as the unit time average power of the power to be supplied.
[0024] The renewable energy power generation device (renewable energy power generation device) 44 is an element that supplies power to the hydrogen production device 20, the compressor 24, etc., with a lower environmental impact than the power grid 40, such as wind power generation or solar power generation. The renewable energy power generation device 40 supplies power to be sold to the hydrogen production device 20, the compressor 24, etc., but may also be configured to be able to supply power to be sold to the power grid 40. However, the renewable energy power generation device 44 does not generally accept purchased power. The renewable energy power generation device 44 provides information such as the supplied unit time average power to the planning device 10.
[0025] The storage battery device 46 is an element that supplements the power supplied to the hydrogen production device 20 and the compression device 24. The storage battery device 46 can be charged by receiving power from the power grid 40 or the renewable energy power generation device 70. The storage battery device 46 may also be in a position to sell power to the power grid 40. The storage battery device 46 provides information such as unit time average power to the planning device 10. As the storage battery device 46 is an auxiliary device, it is not essential and may be omitted.
[0026] (Configuration of planning device) The configuration of the planning device 10 of the embodiment will be described with reference to Fig. 2. The planning device 10 can be realized by a computer including a CPU, a memory, etc. As shown in Fig. 2, the planning device 10 of the embodiment includes an acquisition receiving unit 110, a setting unit 120, a DR supplyable amount searching unit 130, a problem transmitting / receiving unit 140, a plan calculation unit 150, a problem separation unit 160, a resource allocation unit 170, a user interface 180, a display information generating unit 185, and a storage unit 190.
[0027] The acquisition receiving unit 110 is an interface capable of receiving information from the hydrogen production device 20, the storage device 22, the compression device 24, the supply device 26, the transport device 28, and the demand device 30. The acquisition receiving unit 110 acquires information such as the amount of hydrogen per unit time sent from the hydrogen production device 20 to the storage device 22 (production information), the amount of hydrogen stored in the storage device 22 (storage information), the amount of hydrogen per unit time sent from the storage device 22 to the compression device 24, the amount of hydrogen per unit time sent from the compression device 24 to the supply device 26, the amount of hydrogen per unit time sent from the supply device 26 to the transport device 28 (supply information), and information such as the amount of hydrogen per unit time sent from the transport device 28 as a delivery device to the demand device 30 (delivery information).
[0028] The acquisition receiving unit 110 may receive information from the power grid 40, the renewable energy power generation device 44, and the storage battery device 46. The acquisition receiving unit 110 can acquire information (power information) such as the average power per unit time between the power grid 40, the average power per unit time supplied from the renewable energy power generation device 44, the average power per unit time between the storage battery device 46, the average power per unit time to the hydrogen production device 20 and its auxiliaries, and the average power per unit time to the compression device 24.
[0029] The acquisition receiver 110 may further receive information, such as contract information, from the wholesale power trading market 52 and relative power 54. The acquisition receiver 110 may also be configured to acquire information from a forecasting service 60.
[0030] The setting unit 120 is a calculation block that sets parameters and calculation conditions used for optimizing the plan for the entire resource supply chain based on power information, production information, storage information, supply information, delivery information, etc. The setting unit 120 can set information received via the user interface 180 or the acquisition receiving unit 110, for example, as parameters for the plan calculation. The set parameters are stored in the storage unit 190.
[0031] The DR supplyable amount searching unit 130 is a calculation block that searches for a room for procurement by demand response with respect to power procurement from the power grid 40, which is a power procurement source, the wholesale power trading market 52, and the relative power 54. The DR supplyable amount searching unit 130 may be omitted.
[0032] The problem transmission / reception unit 140 is an interface capable of transmitting parameters obtained as a result of planning to each element of the energy system 1. In addition, the problem transmission / reception unit 140 can transmit predetermined arithmetic expressions, parameters, and the like to the computer resource 70 so that the computer resource can process the processes, etc., divided during planning.
[0033] The plan calculation unit 150 is a calculation device that creates a plan related to hydrogen supply (use). The plan calculation unit 150 creates a plan related to hydrogen use by a method such as mathematical optimization using information sent from the hydrogen production device 20, the storage device 22, the compression device 24, the supply device 26, the transport device 28, the demand device 30, the power system 40, the renewable energy power generation device 44, the storage battery device 46, etc. Examples of optimization methods include nonlinear programming and metaheuristic methods (such as genetic algorithms). The plan calculation unit 150 may estimate the mutual power flow between the power procurement source and the power demand destination based on the procurement performance value of each power procurement source and the consumption or storage performance value of each power demand destination. The plan calculation unit 150 may also set an initial value of the environmental load level based on the estimation result of the power flow amount.
[0034] The plan for the entire resource supply chain generated and optimized by the plan calculation unit 150 may be, for example, an overall plan including a production plan in which the hydrogen production device 20 produces hydrogen by receiving power supply from a power procurement source, a storage plan in which the hydrogen produced by the hydrogen production device 20 is stored in the storage device 22, a supply plan in which hydrogen is supplied by the compression device 24 and the supply device 26, and a delivery plan in which hydrogen is delivered by the transport device 28. In other words, the plan (problem) for the entire resource supply chain may include at least two or more problems: the production cost of hydrogen as a fuel material, the supply cost of hydrogen as a fuel material, the environmental impact during the production of hydrogen as a fuel material, and the environmental impact during the supply of hydrogen as a fuel material.
[0035] The problem dividing unit 160 is a calculation block that divides and processes the problem to be calculated by the plan calculation unit 150 according to a predetermined condition. That is, the problem dividing unit 160 divides the entire plan into two or more problem groups according to conditions exemplified by plan units and time series units such as a manufacturing plan, a storage plan, a supply plan, and a delivery plan that constitute an overall plan. The problems divided by the problem dividing unit 160 are transmitted to the computer resource 70 via the problem transmission / reception unit 140 for optimization calculation, and a calculation result is received from the computer resource 70 via the problem transmission / reception unit 140, thereby distributing a part of the calculation of the plan calculation unit 150.
[0036] The resource allocation unit 170 is a calculation block that allocates computer resources according to the characteristics of each problem divided by the problem division unit 160. The resource allocation unit 170 allocates computer resources suitable for the divided problems in terms of, for example, (1) calculation throughput, (2) amount of data that can be communicated, (3) difficulty of the problem that can be calculated, (4) data communication throughput, (5) computer usage fee, and (6) problem format suitable for calculation. The resource allocation unit 170 may perform the allocation from among the computer resources 70 connected to the planning device 10, or may perform the allocation from among the plan calculation unit 150 and the computer resources 70.
[0037] The user interface 180 is an interface for inputting information to the planning device 10, or for providing information created by the planning device 10 to a user via a display device (not shown) or the like. Examples of the user interface 180 include a keyboard, a display device, a mouse, and a touch panel. The user interface 180 can provide the user with the calculation results of the plan calculation unit 150 via a display device or the like.
[0038] The display information generating unit 185 is a calculation block that generates image data (screen display information) for displaying the calculation results of the plan calculation unit 150 on the user interface 180 as a display device. The display information generating unit 185 can generate image data by applying the calculation results of the plan calculation unit 150 to a template stored in the storage unit 190, for example. Examples of the screen display information generated by the display information generating unit 185 include an overall plan, power information, manufacturing information, storage information, supply information, delivery information, forecast information, division results by the problem division unit 160, and allocation results by the resource allocation unit 170.
[0039] The storage unit 190 stores the information acquired by the acquisition receiving unit 110 and stores the calculation results of the plan calculation unit 150. The storage unit 190 may be in the form of a general storage medium or may have a database structure. The storage unit 190 is realized by, for example, a non-volatile memory, a hard disk, a memory disk, or the like.
[0040] (Planning Problems and Parameters) With reference to FIG. 3, a problem to be calculated by the planning device 10 of the embodiment when formulating a plan will be described. FIG. 3 shows examples of independent variables in problems for each element of the energy system 1. The setting unit 120 of the planning device 10 sets an individual independent variable for each combination of input and output of each element. The setting unit 120 uses these independent variables to define input and output variables for each device, and performs mathematical optimization using these variables, making it possible to obtain the power path history between power generation locations and power consumption locations, and to realize optimization of power supply related to hydrogen utilization. In other words, it becomes possible to formulate a plan. Below, the independent variables exemplified in FIG. 3 will be described.
[0041] The independent and dependent variables related to grid power, renewable energy, and batteries are shown below. x market_SELL (t): Average power per unit time [kW] of electricity sold to the wholesale electricity trading market 52 at time t x market_PCHS (t): Average power per unit time received from the wholesale power trading market 52 at time t [kW] x020_SELL (t): Unit time average power [kW] of power sold to relative power 54 at time t x 020_PCHS (t): Average power per unit time [kW] received from relative power 54 at time t x GR_PCHS (t): Average power per unit time received from the power grid 40 at time t [kW] x GR_SELL (t): Average power per unit time [kW] of power sold to the power grid 40 at time t x PV (t): Average power per unit time [kW] of power supplied from the renewable energy power generation device 44 at time t x BAT_DG (t): Average power per unit time [kW] of power supplied from the storage battery device 46 at time t x BAT_CG (t): Average power per unit time [kW] of power charged to the storage battery device 46 at time t x BAT_SOC (t): SOC (State of Charge) [kWh] of the storage battery device 46 at time t (SOC is generally expressed in units of %, but here it is expressed in kWh)
[0042] The independent variables and dependent variables related to the hydrogen production device 20 are shown below. x EC (t): Average power per unit time [kW] of power supplied to the hydrogen production device 20 at time t x EC_AUX (t): Average power per unit time [kW] of power supplied to the auxiliary equipment of the hydrogen production device 20 at time t x EC_OUT (t): The amount of hydrogen per unit time [Nm 3 ]
[0043] The independent and dependent variables for storage device 22 are given below. x BUF_TNK (t): The amount of hydrogen stored in the storage device 22 at time t [Nm 3 ] x BUF_TNK_OUT(t): The amount of hydrogen discharged from the storage device 22 per unit time at time t [Nm 3 ]
[0044] The independent and dependent variables for the compressor 24 are as follows: x CMP_AUX (t): unit time average power [kW] of power supplied to the compressor 24 at time t x CMP_OUT (t): The amount of hydrogen discharged from the compressor 24 per unit time [Nm 3 ]
[0045] The independent and dependent variables for the transport device 28 are shown below. x SHIP (m, t): The amount of hydrogen per unit time [Nm 3 ] y SHIP_TO_DEP (d, m, t): A variable that is 1 if transportation to the demand equipment 30d is performed by the transportation equipment 24m at time t, and is 0 otherwise.
[0046] The acquisition receiver 110 acquires the following information from the demand device 30. Note that the forecasted values may be acquired from a forecasting service 60. The probability distribution or confidence interval of this information becomes a problem for planning the entire resource supply chain. D DEM (d,t): Demand forecast value [Nm 3 ] η RET_TIME (d,t): The predicted time [min] from when the transportation equipment m finishes supplying or refilling to the demand equipment d at time t until it returns to the supply equipment and waits for refilling from the supply equipment. η H2 (d,t): Predicted unit price of hydrogen sold to demand device d at time t [yen / Nm 3 ]
[0047] The acquisition receiver 110 acquires the following information from the wholesale power trading market 52 and the relative power 54. Note that the forecast values may be acquired from a forecast service 60. η MAX_MARKET_PCHS(t): Maximum predicted power purchase value per unit time average power from the wholesale power trading market at time t [kW] η MAX_MARKET_SELL (t): Maximum electricity that can be sold to the wholesale electricity market per unit time average electricity at time t [kW] η MARKET_PCHS_UNIT_PRICE (t): Predicted price of electricity purchased per unit time from the wholesale electricity market at time t [yen / kW] η MARKET_SELL_UNIT_PRICE (t): Estimated selling price per unit of electricity per unit of time to the wholesale electricity trading market at time t [yen / kW]
[0048] (Information to be set in advance) The setting unit 120 sets the following parameters in the storage unit 19. These parameters may be stored in the storage unit 19 via the user interface 180. T: Maximum time to plan (maximum number of planning unit times) T UNIT :Planning unit time [min] (assuming 60 min or less) I: Maximum number of DR products to be planned S PV (t): Average power forecast value per unit time available for generation at time t [kW] T max_past : Maximum number of past times to calculate the baseline M: Maximum number of transport devices x BAT_SOC_UPPER_ASMP (t): SOC allowable upper limit [kWh] of the storage battery device 46 when dividing the problem at time t x BAT_SOC_LOWER_ASMP (t): The lower allowable SOC limit [kWh] of the storage battery device 46 when dividing the problem at time t x BUF_TNK_UPPER_ASMP (t): The upper limit of the hydrogen storage capacity [Nm 3 ] x BUF_TNK_LOWER_ASMP (t): The lower limit of the allowable amount of hydrogen stored in the storage device 22 when dividing the problem at time t [Nm 3 ]
[0049] (Example of power constraint equation) When the plan calculation unit 150 performs optimization of the evaluation function, the setting unit 120 sets a constraint equation, as exemplified below, based on the information acquired by the acquisition and reception unit 110 from each of the renewable energy power generation device 44, the storage battery device 46, and the hydrogen production device 20. The setting equation is stored in the memory unit 19.
[0050]
number
[0051]
number
[0052]
number
[0053]
number
[0054] (Example of a constraint equation for hydrogen) When the plan calculation unit 15 performs optimization of the evaluation function, the plan calculation unit 15 sets the following constraint equations based on the information acquired by the hydrogen information acquisition unit 14. The set equations are stored in the storage unit 19. In the following equations, t = 1, ..., T, and m = 1, ..., M.
[0055]
number
[0056]
number
[0057]
number
[0058]
number
[0059]
number
[0060] (Evaluation function for planning by the planning calculation unit) The plan calculation unit 150 performs optimization of the evaluation function using the constraint equations exemplified above, and calculates an optimal solution. The evaluation function may be cost, greenhouse gas emissions, or a weighted sum of these. In this case, the evaluation function is minimized by optimization. Examples of optimization methods include mixed integer linear programming and metaheuristic methods (such as genetic algorithms).
[0061] (Overall operation of the planning device) The overall operation of the planning device 10 will be described with reference to Fig. 4. The acquisition and reception unit 110 acquires each element of the energy system 1, various predicted values acquired from the prediction service 60, and parameters of each element in the past stored in the storage unit 190, and stores them in the storage unit 190 in preparation for planning (S200).
[0062] The setting unit 120 sets the conditions at the time of planning using the parameters of each element stored in the storage unit 190, various predicted values, and past parameters of each element (S210). The conditions at the time of planning can include initial conditions, end conditions, goals, etc.
[0063] The plan calculation unit 150 and the computer resource 70 perform optimization for a predetermined evaluation function using a constraint equation based on the conditions set by the setting unit 120, and calculate an optimal solution. The calculation result is stored in the storage unit 190 (S220). At this time, the display information generation unit 185 may generate image data using the calculation result stored in the storage unit 190, and provide it to the user through the user interface 180. The DR supplyable amount search unit 130 may execute a search process to determine whether or not it is possible to supply for demand response (S230).
[0064] At this stage, if the specified end time has passed (Yes in S240), the process ends. If the specified end time has not passed (No in S240), or if the specified planning time has not been reached (No in S250), the process waits as is (S240, S250). If the specified planning time has been reached (Yes in S250), the process is repeated from the information acquisition process by the acquisition receiving unit 110 (S200 to S230).
[0065] (Planning in Action) Next, the planning operation of the embodiment in step S220 in Fig. 4 will be described with reference to Fig. 5. Here, the explanation will be given assuming that the elements of the entire resource supply chain are divided into two problem groups.
[0066] When optimizing the plan for the entire resource supply chain by distributing (dividing), there are two possible methods: dividing the elements that make up the resource supply chain, and dividing the time range to be planned by time series. The method of dividing elements divides the problem at a specific element in the entire resource supply chain. The method of dividing by time series divides all problems for the elements in the entire resource supply chain at a specific time into an earlier and later part. The number of divisions is arbitrary.
[0067] The method of division is selected based on, for example, the demand forecast value D SHIP A method based on the predicted value D is exemplified. SHIP If the prediction error is large, it is not possible to calculate an efficient plan. Therefore, the prediction value D SHIP It is preferable to use whether the prediction error is less than a predetermined threshold as a criterion, and to divide the elements if the prediction error is less than the threshold, and to divide them in time series if the prediction error is equal to or greater than the threshold.
[0068] The selection of the division method is not limited to the size of the prediction error. It is also possible to learn past divisions, calculation results, and problem contents by machine learning, and use the learning results to find which division method is better to select.
[0069] The problem division unit 160 divides the demand forecast value D SHIP The demand forecast value D is obtained and the forecast error is compared with a predetermined threshold value (S300). SHIP If the prediction error is less than a predetermined threshold (Yes in S301), the problem dividing unit 160 determines the elements to be divided (dividing boundaries) among the elements of the entire resource supply chain (S302). That is, the problem dividing unit 160 selects element division.
[0070] When optimizing the entire resource supply chain plan by distributing (dividing) it becomes an issue how to handle the boundaries between the divided problem groups. In other words, each problem after division needs to be optimized independently. In addition, when optimizing each divided problem group, the boundaries need to be connected.
[0071] For the boundary line of the problem division, equipment whose values can change suddenly and greatly is not preferable. In other words, equipment whose values change continuously over the time axis is preferable, such as the storage device 22 and the storage battery device 46. In the following explanation, the storage device 22 is defined as the boundary line of the problem division.
[0072] Next, the problem dividing unit 160 works backwards from the input side elements of the divided elements, extracts variables of the connected elements, and sets the variables as problem 1. That is, working backwards from the elements from the storage device 22 to the hydrogen production device 20, the renewable energy power generation device 44, the storage battery device 46, and so on, the variables of each element are extracted, and the problem including these variables is determined as problem 1 (S303).
[0073] Next, the problem dividing unit 160 goes down from the element on the output side of the divided element, extracts variables of the connected elements, and sets this as problem 2. That is, going down from the elements such as the storage device 22, the compression device 24, and the supply device 26, it extracts variables of each element, and determines the problem including these variables as problem 2 (S304).
[0074] Next, the problem division unit 160 adds variables of division elements to each of the problems 1 and 2 (S305). That is, the variable (x BUF_TNK_OUT (t)) is added. This incorporates the common variable of storage device 22 into each of Problem 1 and Problem 2.
[0075] Furthermore, the problem dividing unit 160 divides the divided elements into a predicted value D SHIP Variables up to (X BUF_TNK_OUT , X CMP_OUT , X SHIP ) and a fixed value D SHIP is added (S306).
[0076] The problem dividing unit 160 assumes and fixes values for variables that intersect problem 1 and problem 2 (S307). That is, the value of the variable of the storage device 22 that corresponds to the boundary between problem 1 and problem 2 is fixed to a predetermined value.
[0077] However, if we try to make the boundary line of the divided problem completely coincide with a certain value, inconsistencies may occur during optimization calculations, making it impossible to obtain a solution. In order to avoid this, it is effective to set an allowable width for the boundary line that divides the problem.
[0078] For example, the predicted value D SHIP Based on the predicted value D, target upper and lower limits for the storage device 22 may be set. SHIP When the predicted value D is large, it is expected that the storage capacity of the storage device 22 is often used up, so the upper and lower limits of the storage capacity can be set to low values. SHIP If there is no output X from the storage device 22 BUF_TNK_OUT may be assumed as a hypothesis for the predicted value.
[0079] As a suitable boundary line, the allowable range of parameters for the hydrogen storage device 22 and the storage battery device 46 is the upper limit X BUF_TNK_UPPER_ASMP (t) and lower limit X BUF_TNK_LOWER_ASMP (t), the upper limit X of the remaining capacity of the storage battery device 46 at each timeBAT_SOC_UPPER_ASMP (t) and lower limit X BAT_SOC_LOWER_ASMP (t) and the like may be set. The problem dividing section 160 may incorporate these tolerances into the problem 1 and the problem 2.
[0080] Demand forecast value D SHIP If the prediction error is equal to or greater than a predetermined threshold (No in S301), the problem division unit 160 determines the time to divide the entire operation of the resource supply chain (S308). That is, the problem division unit 160 selects the time division. The division time may be determined by a human being, or the past division times, calculation results, and problem contents may be learned by machine learning, and the division time may be selected based on the learning results.
[0081] Next, the variables of the entire resource supply chain at the time before the division time are extracted and determined as problem 1 (S309).
[0082] Next, variables of the entire resource supply chain at a time after the division time are extracted and set as problem 2 (S310).
[0083] Next, the resource allocation unit 170 allocates computer resources for calculation to each of the problem 1 and the problem 2 divided by the problem division unit 160 (S311). The computer resources to be allocated are selected from the plan calculation unit 150 and the computer resources 70.
[0084] After the allocation to each divided problem is performed, the problem transmitting / receiving unit 140 transmits the problem to be calculated to the allocated plan calculation unit 150 and computer resource 70 (S320). When transmitting the problem to be calculated, the problem transmitting / receiving unit 140 may also transmit necessary parameters.
[0085] The plan calculation unit 150 and the computer resource 70 that have received the problem to be calculated perform optimization of the evaluation function using a predetermined constraint equation to calculate an optimal solution (step S325). The plan calculation unit 150 and the computer resource 70 transmit the calculation result to the resource allocation unit 170.
[0086] The problem transmitting / receiving unit 140 receives the calculation results from the plan calculating unit 150 and the computer resource 70 (step S330).
[0087] The resource allocation unit 170 merges (combines) the calculation results for Problem 1 and Problem 2 received from the plan calculation unit 150 and the computer resource 70, respectively, to calculate an optimal solution for the entire plan generation, and outputs the solution to the storage unit 190 (step S340).
[0088] If the calculated optimal solution contains an error (No in step S350), the process is repeated from problem division (steps S300 to S340). If there is no problem with the calculation result, the display information generating unit 185 outputs the optimal solution stored in the storage unit 190 to the user interface 180.
[0089] In this way, the planning device 10 of the embodiment divides the calculation of planning, distributes it to multiple computer resources, calculates optimal solutions in parallel, and merges the obtained optimal solutions to calculate an optimal solution for the entire planning. Therefore, it is possible to efficiently obtain an optimal solution for the problem of the entire resource supply chain.
[0090] (Method of dividing the problem) Next, a problem division technique in the planning device of the embodiment will be described with reference to FIGS. 6A, 6B, and 7. FIG.
[0091] FIG. 6A is a conceptual diagram of dividing the elements of the entire resource supply chain into a plurality of problems, and FIG. 6B shows the time series changes of each variable when dividing the elements of the entire resource supply chain into a plurality of problems. FIGS. 6A and 6B correspond to the problem division in steps S302 to S306 in FIG. 5. As shown in FIG. 6A, the problem division unit 160 divides the problem into Problem 1 including independent variables in the hydrogen production device 20, storage device 22, power system 40, renewable energy power generation device 44, storage battery device 46, wholesale power trading market 52, and relative power 54 related to hydrogen production, with the storage device 22 of the energy system 1 as a buffer, and Problem 2 including independent variables in the storage device 22, compression device 24, supply device 26, transport devices 28a to 28m, and demand devices 30a to 30d related to the delivery of the produced hydrogen. At this time, the unit time average power (x CMP_AUX (t)) is temporarily fixed to a predetermined value, and then problem 1 and problem 2 are divided and optimized. Also, as shown in FIG. 6B, the problem dividing unit 160 divides the predicted value D SHIP Based on this, the upper and lower limit values (X BUF_TNK_UPPER_ASMP (t), X BUF_TNK_LOWER_ASMP (t)) can be set.
[0092] The example shown in Fig. 7 divides the flow of the resource supply chain into multiple problems in a chronological order, and corresponds to the problem division in steps S307 to S309 in Fig. 5. The problem division unit 160 divides the planning target of the resource supply chain shown in Fig. 3 into problem 1 before a certain time t1 and problem 2 after the time t1. At this time, all elements (variables) are not omitted but are collectively treated as the calculation target.
[0093] Specific methods for performing problem decomposition may include optimization (integer programming, mixed integer programming, nonlinear programming, metaheuristics, etc.), learning (supervised learning, unsupervised learning such as reinforcement learning, deep learning, etc.), etc. Alternatively, learning and optimization may be combined such that only parameter learning is used and the actual problem decomposition is done using optimization.
[0094] (Resource allocation methods) Allocation of resources as computer resources will be described with reference to Fig. 8. Fig. 8 shows an example of criteria by which the resource allocation unit 170 allocates resources.
[0095] In the example shown in Fig. 8, the criteria for resource allocation are (1) computational throughput, (2) the amount of data that can be communicated, (3) the difficulty of the computable problems, (4) data communication throughput, (5) the computer's pay-per-use fee, and (6) the type of problem suitable for computation. In other words, for problems to which resources are to be allocated, the type of resources to be allocated is narrowed down based on the indicators of whether the computational throughput is high, medium, or low, whether the amount of data that can be communicated is high, medium, or low, and whether the difficulty of the computable problems is complete or difficult ("A" in the figure).
[0096] Furthermore, in the indicator A in the figure, the types of resources to be allocated are further narrowed down ("B" in the figure) using indicators such as whether the data communication throughput is high / medium / low, whether the computer's pay-per-use fee is high / medium / low, and whether the problem format is "quadratic mixed integer linear programming," "quadratic mixed integer nonlinear programming," "linear mixed integer linear programming," "linear mixed integer nonlinear programming," or "nonlinear least squares problem." An indicator for further narrowing down may be set ("C" in the figure).
[0097] The resource allocation unit 170 of the embodiment determines where the six indexes correspond in the matrix shown in Fig. 8, and allocates the optimal resource for the combination of indexes. The resource to be allocated can be selected from the plan calculation unit 150, the computer resource 70, etc.
[0098] A specific method for allocating resources may use optimization (integer programming, mixed integer programming, nonlinear programming, metaheuristics, etc.) or learning (supervised learning, unsupervised learning such as reinforcement learning, deep learning, etc.). In other words, the suitability of resources may be determined by an optimization technique, or the optimal resource obtained by learning may be adopted. Similarly, a combination may be used in which only parameters are learned and actual resource allocation is performed using optimization.
[0099] The index used for resource allocation is not limited to the example shown in FIG. 8. For example, the communication delay from the location of the monitoring and control to the location of the actual computer resources may be included in the evaluation. In addition, when performing a simulation using a tool, the appropriateness of the resources may be determined and allocated in consideration of the computational load on the tool resources. For example, for high-load processing such as 3D-CAE, it is conceivable to allocate resources with high processing power such as a quantum computer.
[0100] In addition, in resource allocation, a classical computer may be used to train a classifier, and the results may be sent to a quantum computer. When the resource is a quantum computer, not only the preparation state but also the measurement time may be taken into consideration, as well as the unitary transformation, which is the actual processing. In addition, the processing time for state preparation may be taken into consideration when transferring data to the quantum computer.
[0101] Regarding the delivery plan, since the result is reduced to a problem of only variables that have two values, 0 or 1, the weight of resource allocation to the quantum computer may be increased. In addition, data may be exchanged between distributed computers as computing resources. For example, when transferring data from a classical computer to a quantum computer, the method of state preparation may be specified.
[0102] When the resource is a quantum computer, a probabilistic result may be returned from the quantum computer. In that case, the procedure of sending the problem in step S320, the planned calculation process in step S325, and receiving the result in step S330 in Fig. 5 may be repeated multiple times. Therefore, the resource allocation may be performed by setting an upper limit on the number of times the procedure is repeated, or by estimating the number of times the procedure is repeated in advance and taking the estimated value into consideration.
[0103] It should be noted that user authentication for resources to be distributed may be configurable.
[0104] According to the planning device of this embodiment, since the process is divided with consideration of the planning element for optimizing the entire supply chain of resources, it is possible to realize large-scale planning. In particular, since the production and delivery of resources are divided into separate problems, the optimization calculation can be smoothly executed.
[0105] Furthermore, according to the planning device of the embodiment, resources are allocated according to the speed, bandwidth, nature of the calculations, etc. required to process the divided problems, so that the calculations related to the overall planning can be optimized.
[0106] Second embodiment A planning device according to a second embodiment will be described with reference to Fig. 9. Fig. 9 shows an operation example according to the second embodiment. This embodiment adds learning results to the planning device according to the first embodiment. In the following description, common elements and operations are denoted by common reference symbols, and duplicated descriptions will be omitted.
[0107] The acquisition and reception unit 110 acquires each element of the energy system 1, various predicted values acquired from the prediction service 60, and parameters of each element in the past stored in the memory unit 190, and stores them in the memory unit 190 in preparation for planning (S200).
[0108] Here, if the planning process of the planning device 10 is not the first time (No in step S202), the resource allocation unit 170 executes a learning calculation using the calculation result of the previous plan stored in the storage unit 190 and each element, various predicted values, and parameters of each element newly acquired by the acquisition receiving unit 110 (step S204). The learning result is stored in the storage unit 190.
[0109] When the planning process of the planning device 10 is the first time (Yes in step S202) or when the planning calculation unit 150 executes a learning calculation instead of the first time (step S204), the setting unit 120 sets the conditions at the time of planning using the parameters of each element stored in the storage unit 190, various predicted values, and parameters of each element in the past (S210). The conditions at the time of planning can include initial conditions, end conditions, goals, etc.
[0110] The plan calculation unit 150 and the computer resource 70 perform optimization for a predetermined evaluation function using a constraint equation based on the conditions set by the setting unit 120, and calculate an optimal solution. The calculation result is stored in the storage unit 190 (S220). At this time, the display information generation unit 185 may generate image data using the calculation result stored in the storage unit 190, and provide it to the user through the user interface 180. The DR supplyable amount search unit 130 may execute a search process to determine whether or not it is possible to supply for demand response (S230).
[0111] At this stage, if the specified end time has passed (Yes in S240), the process ends. If the specified end time has not passed (No in S240), the problem division unit 140 determines whether or not a new plan needs to be made based on the calculation result and the learning result stored in the memory unit 190 (step S245). For example, if there is no difference between the learned previous (past) planning result and the current planning result, planning can be omitted even if it is the specified planning time.
[0112] As a result of the determination, if planning is to be omitted (No in S255), the process waits as is (S240, S255). If planning is to be executed (Yes in S255), the process is repeated from the information acquisition process by the acquisition receiving unit 110 (S200 to S230).
[0113] According to this embodiment, since the planning results are learned, calculation processing can be omitted when there are few changes in the independent variables, etc., making it possible to conserve computer resources.
[0114] Third embodiment A planning device according to a third embodiment will be described with reference to Fig. 10. Fig. 10 shows an example of operation according to the third embodiment. In this embodiment, a learning result is added to the planning device according to the first embodiment, and the learning result is used for planning calculation. In the following description, common elements and operations are denoted by common reference numerals, and duplicated descriptions will be omitted.
[0115] The acquisition and reception unit 110 acquires each element of the energy system 1, various predicted values acquired from the prediction service 60, and parameters of each element in the past stored in the memory unit 190, and stores them in the memory unit 190 in preparation for planning (S200).
[0116] Here, if the planning process of the planning device 10 is not the first time (No in step S202), the resource allocation unit 170 executes a learning calculation using the calculation result of the previous plan stored in the storage unit 190 and each element, various predicted values, and parameters of each element newly acquired by the acquisition receiving unit 110 (step S204). The learning result is stored in the storage unit 190.
[0117] When the planning process of the planning device 10 is the first time (Yes in step S202) or when the planning calculation unit 150 executes a learning calculation instead of the first time (step S204), the setting unit 120 sets the conditions at the time of planning using the parameters of each element stored in the storage unit 190, various predicted values, parameters of each element in the past, and the learning results thereof (S212). The conditions at the time of planning can include initial conditions, end conditions, goals, etc.
[0118] The plan calculation unit 150 and the computer resource 70 perform optimization for a predetermined evaluation function using a constraint equation based on the conditions set by the setting unit 120, and calculate an optimal solution. The calculation result is stored in the storage unit 190 (S220). At this time, the display information generation unit 185 may generate image data using the calculation result stored in the storage unit 190, and provide it to the user through the user interface 180. The DR supplyable amount search unit 130 may execute a search process to determine whether or not it is possible to supply for demand response (S230).
[0119] At this stage, if the specified end time has passed (Yes in S240), the process ends. If the specified end time has not passed (No in S240), or if the specified planning time has not been reached (No in S250), the process waits as is (S240, S250). If the specified planning time has been reached (Yes in S250), the process is repeated from the information acquisition process by the acquisition receiving unit 110 (S200 to S230).
[0120] According to this embodiment, the planning results are learned, and the planning is performed using the learning results in addition to the acquired predicted values, actual values, and past data, so that it is possible to speed up the calculation process when there are few changes in the independent variables, etc. Also, since the learning results are used, it is possible to realize more accurate planning.
[0121] (Fourth embodiment) Next, the planning operation of the fourth embodiment will be described with reference to Fig. 11. The planning operation of this embodiment is an operation in which, after resource allocation, an operation for correcting a problem in a question format suitable for the allocated resources is added. In the following description, operations common to the first to third embodiments are denoted by common reference symbols, and duplicated explanations will be omitted.
[0122] When the setting unit 120 sets the conditions for planning, the problem dividing unit 160 divides each problem such as the independent variables shown in FIG. SHIPA division method is selected based on the presence or absence of a prediction error and the difference from a threshold value, and the like, and the problem is divided into multiple problems by element or time series (S300 to S310).
[0123] Next, the resource allocation unit 170 allocates computer resources for calculation to each of the problem 1 and the problem 2 divided by the problem division unit 160 (S311). The computer resources to be allocated are selected from the plan calculation unit 150 and the computer resources 70.
[0124] When resources are allocated to each divided problem, the resource allocation unit 170 modifies the format of the allocated problem according to the characteristics of the allocated resources (step S312). For example, if a quantum computer is selected as the computational resource as a result of the resource allocation, it is desirable to modify the problem format to one suitable for the computation characteristics of the quantum computer. The resource allocation unit 170 modifies the divided problem according to the allocated resources based on a problem modification algorithm previously stored in the storage unit 190.
[0125] After the divided problems are assigned to each other and the problems are corrected, the problem transmitting / receiving unit 140 transmits the problem to be calculated to the assigned plan calculation unit 150 and the computer resource 70 (S320). When transmitting the problem to be calculated, the problem transmitting / receiving unit 140 may also transmit necessary parameters.
[0126] The plan calculation unit 150 and the computer resource 70 that have received the problem to be calculated perform optimization of the evaluation function using a predetermined constraint equation to calculate an optimal solution (step S325). The plan calculation unit 150 and the computer resource 70 transmit the calculation result to the resource allocation unit 170.
[0127] The problem transmitting / receiving unit 140 receives the calculation results from the plan calculating unit 150 and the computer resource 70 (step S330).
[0128] The resource allocation unit 170 merges the calculation results received from the plan calculation unit 150 and the computer resource 70 to calculate an optimal solution for the entire plan formulation, and outputs the solution to the storage unit 190 (step S340).
[0129] If the calculated optimal solution contains an error (No in step S350), the process is repeated from problem division (steps S300 to S340). If there is no problem with the calculation result, the display information generating unit 185 outputs the optimal solution stored in the storage unit 190 to the user interface 180.
[0130] In this way, the planning device 10 of the embodiment divides the calculation of planning, distributes it to multiple computer resources, calculates optimal solutions in parallel, and merges the obtained optimal solutions to calculate an optimal solution for the entire planning. Therefore, it is possible to efficiently obtain an optimal solution for the problem of the entire resource supply chain.
[0131] This embodiment also allows for more efficient planning since problems are fixed according to allocated resources.
[0132] Fifth embodiment Next, the planning operation of the fifth embodiment will be described with reference to Fig. 12. The planning operation of this embodiment adds an encryption processing operation before transmitting the problem for which resource allocation has been performed to the computer resource. In the following description, operations common to the first to fourth embodiments are denoted by common reference numerals, and duplicated descriptions will be omitted. When the setting unit 120 sets the conditions for planning, the problem dividing unit 160 divides each problem such as the independent variables shown in FIG. SHIP A division method is selected based on the presence or absence of a prediction error and the difference from a threshold value, and the like, and the problem is divided into multiple problems by element or time series (S300 to S310).
[0133] Next, the resource allocation unit 170 allocates computer resources for calculation to each problem divided by the problem division unit 160 (S311). The computer resources to be allocated are selected from the plan calculation unit 150 and the computer resources .
[0134] After the allocation to each divided problem is performed, the resource allocation unit 170 encrypts the divided problem using a predetermined algorithm (step S314). The encryption method is such that each of the allocated computer resources 70 can decrypt the data. In addition to the divided problem, necessary parameters and calculation conditions may also be encrypted.
[0135] The problem transmitting / receiving unit 140 transmits the encrypted problem to be calculated to the plan calculation unit 150 and the computer resource 70 (S320). When transmitting the problem to be calculated, the problem transmitting / receiving unit 140 may transmit encrypted necessary parameters and calculation conditions together.
[0136] The plan calculation unit 150 and the computer resource 70, which have received the problem to be calculated, use a predetermined constraint equation to perform optimization of the evaluation function and calculate an optimal solution (step S325). The plan calculation unit 150 and the computer resource 70 transmit the calculation result to the problem transmission / reception unit 140. At this time, the plan calculation unit 150 and the computer resource 70 may apply a predetermined encryption to the calculated optimal solution.
[0137] The problem transmitting / receiving unit 140 receives the calculation results from the plan computing unit 150 and the computer resource 70 (step S330). If the received calculation results are encrypted, the resource allocation unit 170 also decrypts them.
[0138] The resource allocation unit 170 merges the calculation results received from the plan calculation unit 150 and the computer resource 70 to calculate an optimal solution for the entire plan formulation, and outputs the solution to the storage unit 190 (step S340).
[0139] If the calculated optimal solution contains an error (No in step S350), the process is repeated from problem division (steps S300 to S340). If there is no problem with the calculation result, the display information generating unit 185 outputs the optimal solution stored in the storage unit 190 to the user interface 180.
[0140] In this way, the planning device 10 of the embodiment divides the calculation of planning, distributes it to multiple computer resources, calculates optimal solutions in parallel, and merges the obtained optimal solutions to calculate the optimal solution for the entire planning. Therefore, it is possible to efficiently obtain the optimal solution for the problem of the entire resource supply chain. In addition, since the planning device 10 of this embodiment encrypts the problem to be sent to the computer resource, security can be ensured when the problem is sent to the computer resource 70 via, for example, the Internet.
[0141] Sixth embodiment Next, the overall operation of the planning device of the sixth embodiment will be described with reference to Fig. 13. The operation example of the planning device shown in Fig. 13 is to execute a search for a DR supplyable amount in addition to the basic plan planning. In the following description, steps common to the overall operation of the first embodiment are denoted by the same reference numerals, and duplicated description will be omitted.
[0142] The acquisition and reception unit 110 acquires each element of the energy system 1, various predicted values acquired from the prediction service 60, and parameters of each element in the past stored in the memory unit 190, and stores them in the memory unit 190 in preparation for planning (S200).
[0143] The setting unit 120 sets the conditions at the time of planning using the parameters of each element stored in the storage unit 190, various predicted values, and past parameters of each element (S210). The conditions at the time of planning can include initial conditions, end conditions, goals, etc.
[0144] The plan calculation unit 150 and the computer resource 70 execute a search process to determine whether or not the energy can be provided for demand response, and also perform optimization of a predetermined evaluation function using a constraint equation based on the conditions set by the setting unit 120 to calculate an optimal solution. The calculation result is stored in the storage unit 190 (S222). At this time, the display information generation unit 185 may generate image data using the calculation result stored in the storage unit 190, and provide the image data to the user via the user interface 180.
[0145] At this stage, if the specified end time has passed (Yes in S240), the process ends. If the specified end time has not passed (No in S240), or if the specified planning time has not been reached (No in S250), the process waits as is (S240, S250). If the specified planning time has been reached (Yes in S250), the process is repeated from the information acquisition process by the acquisition receiving unit 110 (S200 to S230).
[0146] According to the planning device of this embodiment, a search process for determining whether or not supply of power through demand response is possible is performed in conjunction with the calculation of an optimal solution, so that an optimal solution appropriate for receiving power supply through demand response can be obtained.
[0147] Seventh embodiment Next, an application example of problem division in planning will be described with reference to Fig. 14. Fig. 14 is a diagram showing transport devices and demand devices for which planning is performed by the planning device of the seventh embodiment, and is an example in which there are multiple transport device sales offices for multiple transport devices and multiple demand devices.
[0148] As shown in Figure 14, the amount of calculation required to calculate the optimal solution for a large-scale and complex supply and delivery problem is enormous. Here, a transportation equipment office is imagined as, for example, a driver's office (where work begins and ends) or a parking area for trailer heads (trucks).
[0149] In order to calculate an optimal solution for such a large-scale, complex problem, the problem division unit 160 performs one or more of the following: problem division, problem reduction, partial omission of the problem, plan length reduction, and plan length division. In other words, by dividing or omitting the amount of calculation, it becomes possible to regard a quasi-optimal solution to a large-scale, complex problem as an optimal combination result, and it becomes possible to reduce the calculation load of the entire resource supply chain.
[0150] Eighth embodiment Next, a planning device according to an eighth embodiment will be described with reference to Fig. 15 and Fig. 16. The planning device according to the eighth embodiment further includes a prediction unit 175. In the following description, elements and operations common to the first embodiment are denoted by common reference numerals, and duplicated descriptions will be omitted.
[0151] 15, the planning device 11 of this embodiment includes a prediction unit 175. The prediction unit 175 is a calculation block that predicts independent variables and the like necessary for the planning of the plan calculation unit 150 that cannot be acquired by the acquisition and reception unit 110. Examples of data (prediction information) predicted by the prediction unit 175 include at least one of the required power or heat of a demand device that is a consumer of hydrogen as a fuel material, the consumption amount of hydrogen as a fuel material, the time when the transportation device arrives at a supply device at a base station, and the amount of hydrogen supply that the transportation device receives from the supply device at the base station.
[0152] Next, the overall operation of the planning device 11 will be described with reference to Fig. 16. The acquisition receiving unit 110 acquires each element of the energy system 1, various predicted values acquired from the prediction service 60, and parameters of each element in the past stored in the storage unit 190, and stores them in the storage unit 190 in preparation for planning (S200).
[0153] The prediction unit 175 predicts and calculates data required for the plan calculation unit 150 and the computer resource 70 to make a plan, using the parameters and the like acquired by the acquisition and reception unit 110 (S206).
[0154] The setting unit 120 sets the conditions at the time of planning (S210) using the parameters of each element stored in the storage unit 190, various predicted values, parameters of each element in the past, and data predicted by the prediction unit 175. The conditions at the time of planning can include initial conditions, end conditions, goals, and the like.
[0155] The plan calculation unit 150 and the computer resource 70 perform optimization for a predetermined evaluation function using a constraint equation based on the conditions set by the setting unit 120, and calculate an optimal solution. The calculation result is stored in the storage unit 190 (S220). At this time, the display information generation unit 185 may generate image data using the calculation result stored in the storage unit 190, and provide it to the user through the user interface 180. The DR supplyable amount search unit 130 may execute a search process to determine whether or not it is possible to supply for demand response (S230).
[0156] At this stage, if the specified end time has passed (Yes in S240), the process ends. If the specified end time has not passed (No in S240), or if the specified planning time has not been reached (No in S250), the process waits as is (S240, S250). If the specified planning time has been reached (Yes in S250), the process is repeated from the information acquisition process by the acquisition receiving unit 110 (S200 to S230).
[0157] According to the planning device of this embodiment, it is possible to predict data required for planning and to supplement missing information.
[0158] Ninth embodiment Screen display information generated by the planning device of the ninth embodiment will be described with reference to Fig. 17 to Fig. 23. The display information generating unit 185 of the planning devices 10 and 11 of the embodiment generates screen display information that can be displayed on the user interface 180 as a display device for the optimal solution calculated by the plan calculation unit 150 and the computer resource 70. The screen display information is generated by, for example, applying data indicating the optimal solution to a template as a model stored in the storage unit 190.
[0159] The display information generating unit 185 may generate screen display information indicating the division of the problem by the problem dividing unit 160 and the state of resource allocation by the resource allocating unit 170. Fig. 17 to Fig. 19 show examples of screen display information indicating the state of resource allocation by the resource allocating unit 170.
[0160] The screen display information shown in Fig. 17 is an example of information that displays a status display screen for each resource. The information display screen associates resources A to G as computer resources with the average throughput performance, reliability, the presence or absence of malfunctions, the type of assigned problem, parameters of the basis for assignment, and the like for each resource. In the status display screen, resources that have not been assigned are shown with vertical parallel lines. The display with parallel lines may be shown in a color such as gray.
[0161] The average throughput performance on the status display screen shows the amount of throughput delay for each resource. In the example shown in Fig. 17, for a resource, a large delay compared to the expected throughput value is shown by parallel lines inclined counterclockwise, a continuing delay is shown by parallel lines inclined counterclockwise and clockwise that intersect with each other, and a clockwise parallel line inclined as expected is shown. The display by the parallel lines may be shown in a color such as red, yellow, or light blue.
[0162] The reliability indicates the reliability of each resource, and may be actual or predicted reliability, or a mixture of both. In the example shown in Fig. 17, the reliability as expected is shown by parallel lines inclined clockwise, and the reliability below the expected value is shown by parallel lines inclined counterclockwise. The display by the parallel lines may be shown in color, for example, light blue, red, etc.
[0163] The presence or absence of a malfunction indicates the presence or absence of a malfunction for each resource. In the example shown in Fig. 17, when no malfunction occurs, it is shown by parallel lines tilted clockwise, and when a malfunction occurs, it is shown by parallel lines tilted counterclockwise. The display by the parallel lines may be shown in color, such as light blue or red.
[0164] The allocation problem shows the type of problem that is divided and assigned to each resource. In the example shown in Fig. 17, a manufacturing planning problem using linear mixed integers is assigned to resource A, a delivery planning problem using linear mixed integers is assigned to resource B, a delivery planning problem using linear mixed integers is assigned to resource D, and a power generation forecasting problem is assigned to resource F. There is no allocation to resources C, E, and G.
[0165] When a button for displaying assigned problems is selected via a user interface 180 serving as an input device such as a mouse or a touch panel, display screen information showing the problems assigned to the resource is displayed as a status display screen. Fig. 18 shows an example of display screen information showing the problems assigned to the resource.
[0166] In the example shown in Fig. 18, the evaluation formula minE, the constraint formulas p(t), q(t), a variable explanation button, a fault analysis button, and the like are displayed. When the variable explanation button is selected through the user interface 180, an explanation of the variables used in the problem is displayed. Fig. 19 is a diagram showing an example of the variable explanation. In the example shown in Fig. 19, it is shown that the variable E indicates the power required for production in units of 30-minute average kW, the variable p(t) indicates the power purchased from the wholesale power trading market in units of 30-minute average kW, and the variable q(t) indicates the predicted power generation value of the on-site renewable energy power generation device in units of 30-minute average kW.
[0167] When a defect analysis button for a problem where a defect has occurred is selected through the user interface 180, the defect is analyzed by the planning device of the present invention or by some other resource, and the location that is estimated to be defective is presented as a result. Fig. 20 is a diagram showing an example of the defect analysis result. In the example shown in Fig. 20, it is shown that a contradiction has occurred in the constraint equations p(t) and q(t).
[0168] The screen display information showing the status display screen for each resource is configured so that subtitles are displayed when the average throughput performance, reliability, and allocation basis parameters are selected through the user interface 180. Fig. 21 shows a state in which the average throughput performance is selected and the computation throughput and data communication throughput are displayed as subtitles. Fig. 22 shows a state in which the reliability is selected and the mean time to recovery (MTTR), mean time between failures (MTBF), availability, and failure rate curve are displayed as subtitles. Fig. 23 shows a state in which the allocation basis parameters are selected and the computation throughput, communicable data volume, difficulty of computable problems, data communication throughput, and computer usage fee are displayed as subtitles.
[0169] In this way, according to the computing device of this embodiment, the state of resource allocation by the resource allocation unit 170 can be presented to the user, so that the user can manually optimize the resource allocation via the user interface 180.
[0170] Although several embodiments of the present invention have been described above, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope and spirit of the invention, and are included in the scope of the invention and its equivalents described in the claims. [Explanation of symbols]
[0171] 1...energy system, 10, 11...planning device, 110...acquisition and reception unit, 120...setting unit, 130...DR supply available amount search unit, 140...transmission unit, 150...planning calculation unit, 160...problem division unit, 170...resource allocation unit, 180...user interface, 185...display information generation unit, 190...memory unit, 20...hydrogen production device, 22...storage device, 24...compression device, 26...supply device, 28...transportation device, 30...demand device, 40...power system, 44...renewable energy power generation device (renewable energy power generation device), 46...storage battery device, 52...wholesale electricity trading market, 54...relative power, 60...forecasting service, 70...computer resources.
Claims
1. connected to a power source, a fuel material production device, a fuel material storage device, a fuel material supply device, and a fuel material delivery device; A planning device that optimizes an overall plan including a production plan in which the fuel material production device produces a fuel material by receiving power supply from the power procurement source, a storage plan in which the fuel material storage device stores the fuel material, a supply plan in which the fuel material supply device supplies the fuel material, and a delivery plan in which the fuel material delivery device delivers the fuel material, an information receiving unit that receives power information including the amount of power supply from the power procurement source, production information including the amount of production in the fuel material production device, storage information including the amount stored in the fuel material storage device, supply information including the amount of supply from the fuel material supply device, and delivery information including the amount of delivery from the fuel material delivery device; a setting unit that sets calculation conditions for optimizing the overall plan based on the power information, the production information, the storage information, the supply information, and the delivery information; a problem dividing unit that divides problems related to the manufacturing plan, the storage plan, the supply plan, and the delivery plan, which constitute the overall plan, into two or more problem groups based on a predetermined condition; a resource allocation unit that allocates two or more computer resources having different characteristics to each of the divided problem groups according to the characteristics of the resources; a problem transmission / reception unit that transmits the problem set to which the computer resource is assigned and the calculation conditions to the computer resource and receives calculation results from the computer resource; a combining unit that combines the received calculation results for each problem group to generate the overall plan; A planning device comprising:
2. The planning device according to claim 1, characterized in that the problem divided by the problem division unit is at least two or more problems: the production cost of the fuel material, the supply cost of the fuel material, the environmental impact level during the production of the fuel material, and the environmental impact level during the supply of the fuel material.
3. The planning device according to claim 1 , further comprising a demand response available supply amount searching unit that searches for a margin for power procurement by demand response from the power procurement source for the overall plan.
4. A storage unit that stores the overall plan generated in the past; 2. The planning device according to claim 1, further comprising a plan calculation unit that determines whether or not a next overall plan needs to be created based on the generated overall plan and the overall plans generated in the past.
5. A storage unit that stores the overall plan generated in the past and a learning result based on the overall plan, The planning device according to claim 1, characterized in that the setting unit sets calculation conditions based on the learning results, including one or more of the following: a plan length, an assumed upper limit value of the remaining amount for each time of the storage device, an assumed lower limit value of the remaining amount for each time of the storage device, and an assumed fixed value for each time of the auxiliary device.
6. 2. The planning device according to claim 1, wherein said resource allocation unit, after allocating said computer resources to each of said divided problem groups, modifies said problem groups into a format suited to said allocated computer resources.
7. 2. The planning device according to claim 1, further comprising an encryption unit that encrypts the group of problems to which computer resources are allocated and the operation conditions transmitted by the problem transmission unit.
8. The planning device described in claim 1, characterized in that the problem division unit performs one or more of problem division, problem reduction, omission of part of the problem, plan length shortening, and plan length division for the overall plan, including the plan for supplying the fuel material manufactured by the fuel material manufacturing device by the fuel material storage device and the fuel material supply device, and the plan for delivering the fuel material by the fuel material delivery device.
9. The planning device according to claim 1, characterized in that the problems divided by the problem division unit include problems relating to the probability distribution or confidence interval of at least one of the maximum purchaseable power prediction from the wholesale power trading market, the maximum sellable power prediction to the wholesale power trading market, the average purchase price prediction from the wholesale power trading market, the average sale price prediction to the wholesale power trading market, and the fuel material demand prediction value of fuel material consumers.
10. The planning device according to claim 1, characterized in that the problems divided by the problem division unit further include problems relating to a probability distribution or confidence interval for at least one of a renewable energy power generation prediction and a predicted time until a transport device for a fuel material consumer returns to a supply device and becomes ready for refilling.
11. 2. The planning device according to claim 1, wherein the problem divided by said problem dividing unit further includes a problem relating to an assumed lower limit value and an assumed upper limit value of a remaining battery charge for each time period.
12. 2. The planning device according to claim 1, further comprising a demand response available supply search unit that searches for room for power procurement through demand response from the power procurement source for problems related to the power information, the production information, and the storage information.
13. A prediction unit that generates prediction information including at least one of the power or heat required by a demand device that is a demand device of the fuel material, the consumption amount of the fuel material, the time when the transportation device arrives at a supply device at a base station, and the amount of hydrogen supplied to the transportation device from the supply device at the base station, 2. The planning device according to claim 1, wherein the setting unit further sets a calculation condition for formulating the overall plan based on the forecast information.
14. The planning device according to any one of claims 1 to 13, further comprising a display information generation unit that generates screen display information capable of displaying on a display device at least one of the overall plan, the power information, the manufacturing information, the storage information, the supply information, the delivery information, the forecast information, the division result by the problem division unit, and the allocation result by the resource allocation unit.
Citation Information
Patent Citations
Distributed power supply system and method of controlling the same
JP2011002929A
Optimization device and optimization program for gas energy supply-demand system
JP2019144897A
Distributed computing systems and methods
JP2015517144A
Planning device, method for generating operation plan, method for producing hydrogen, and program
WO2020196889A1
