Waste traceability system and waste traceability method

Abstract

To conduct tracing and simulation in a distributed and parallel manner, thereby facilitating handling of a huge number of inputs and parameters.SOLUTION: A waste traceability system includes: tracing means that traces a waste amount generated by each of a plurality of related business operators; simulation means that performs simulation of the waste amount after varying raw material input quantities and parameters; execution environment allocation means that allocates an execution environment to the tracing means and issues an identification number for identifying the tracing means, and also allocates an execution environment for performing the simulation to the simulation means and issues an identification number for identifying the simulation means; a plurality of partial model servers that calculate the waste amount for each business operator based on the raw material input quantities and the parameters; and an integration server that integrates calculation results from the plurality of partial model servers.SELECTED DRAWING: Figure 1A

Description

[Technical Field]

[0001] The present invention relates to a waste tracing system and a waste tracing method. [Background technology]

[0002] In recent years, as a measure to combat global warming, there has been a need to understand greenhouse gas (GHG) emissions. It is difficult to directly measure carbon dioxide (CO2) or methane (CH4) emissions into the atmosphere using inexpensive sensors without relying on large-scale exhaust measurement equipment. CO2 or CH4 emissions are typically estimated using information from concentration sensors and the reaction efficiency design values ​​of various reactors. Furthermore, to estimate GHG emissions in a given region or related industry, it is necessary to consider the activities of all involved parties: producers, consumers, recyclers, and final processors.

[0003] One method proposed to address this need is digital twins, which model real-world behavior and perform simulations in a virtual space. This makes it possible to trace physical quantities that cannot be directly grasped through computational processing in a virtual space. Tracing involves understanding and recording when, where, and how much of a substance that impacts the environment was used. Digital twins also make it possible to virtually change parameters or input quantities and perform simulations. This makes it possible to make decisions regarding GHG emission reductions by evaluating, for example, the effect of improving reaction efficiency when installing new equipment, or the GHG emission reductions relative to fuel costs when increasing the amount of recycled fuel purchased.

[0004] The invention of Patent Document 1 applies a digital twin (referred to in the document as an "energy virtual twin") to a power grid system to obtain a detailed understanding of a wide range of energy demand and consumption. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Publication No. 2023-163160 Summary of the Invention [Problem to be solved by the invention]

[0006] Generally, waste tracing systems for GHGs, chemical substances, etc. are huge systems that involve all sectors of industry, and are expensive to build. Therefore, it is difficult for a waste tracing system provider to provide a digital twin of society as a whole on its own, and collaboration between multiple industries and multiple business entities is necessary. In other words, in order for a waste tracing system provider to build a digital twin of society as a whole, it will need to receive know-how from related businesses. Therefore, the possession of information that related businesses wish to keep confidential will become an obstacle to building a waste tracing system. The following (1) and (2) are expected to be challenges for such systems.

[0007] (1) In simulations, parameters must be set for multiple different conditions, and a mechanism is needed to run the simulation in parallel with daily tracing. (2) Waste tracing spans all industrial sectors, and the number of parameters to be handled is enormous. It is difficult for businesses to grasp all the calculation parameters. However, the invention of Patent Document 1 does not disclose a specific configuration that satisfies the above (1) and (2). Therefore, an object of the present invention is to perform tracing and simulation in parallel in a distributed manner, thereby facilitating the handling of a huge amount of input and parameters. [Means for solving the problem]

[0008] The waste tracing system of the present invention is characterized by having: tracing means for tracing the amount of waste for each of a plurality of related businesses; simulation means for simulating the amount of waste after changing the amount of raw material input and parameters; execution environment allocation means for assigning an execution environment for performing the tracing to the tracing means and issuing an identification number to identify the tracing means, and for assigning an execution environment for performing the simulation to the simulation means and issuing an identification number to identify the simulation means; a plurality of partial model servers for calculating the amount of waste for each of the businesses based on the amount of raw material input and the parameters; and an integration server for integrating the calculation results of the plurality of partial model servers. Other means will be described in the detailed description of the invention. [Effects of the Invention]

[0009] According to the present invention, tracing and simulation are performed in parallel in a distributed manner, making it possible to easily handle a huge number of inputs and parameters.

[0010] The present invention is particularly suitable for cases where the amount of a substance that places a burden on the environment cannot be directly measured and the amount of the substance that is produced is evaluated using a theoretical calculation or an estimation model based on an approximate formula. The present invention can be applied to, for example, tracing of GHG emissions, life cycle management of chemical substances, tracing of recycled PET (Polyethylene Terephthalate) bottles, and the like.

[0011] The present invention also assumes a distributed system that facilitates collaboration between multiple industries. Each related business has its own server, and each server provides only its computing power to the waste tracing system. This configuration prevents the calculation know-how of related businesses from leaking out. Furthermore, by predicting emissions through simulation in addition to tracing, the present invention makes it possible to use the same model to simultaneously propose equipment replacement and reduce emissions through operational improvements. This added value alleviates the perceived high prices that customers feel for waste tracing systems, which tend to be large systems. [Brief explanation of the drawings]

[0012] [Figure 1A] 1 is a configuration example of a waste tracing system. [Figure 1B] 10 is a configuration example of a partial model server. [Figure 2] 1 is an example of an entire model template. [Figure 3] FIG. 10 is a diagram illustrating an input / output relationship between an entire model template and a partial model. [Figure 4A] This is the data flow between the components in the first half of the trace operation. [Figure 4B] This is the flow of data between the components in the latter half of the trace operation. [Figure 5A] 10 is a flowchart of the first half of a tracing operation. [Figure 5B] 10 is a flowchart of the second half of the tracing operation. [Figure 6] FIG. 10 is a diagram illustrating an example of a sensor / model correspondence table. [Figure 7A] This shows the data flow between the components in the first half of the simulation operation. [Figure 7B] This shows the data flow between the components in the latter half of the simulation operation. [Figure 8A] 10 is a flowchart of the first half of the simulation operation. [Figure 8B] 10 is a flowchart of the second half of the simulation operation. [Figure 9] FIG. 10 is a diagram illustrating an example of a partial model server table. [Figure 10] FIG. 10 illustrates a variation of the waste tracing system. [Figure 11] FIG. 10 is a diagram illustrating the operation of an execution process monitoring means. [Figure 12] FIG. 10 is a diagram illustrating the operation of the model validity monitoring means. [Figure 13] FIG. 1 is a hardware configuration diagram of the waste tracing system. DETAILED DESCRIPTION OF THE INVENTION

[0013] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Fig. 1A shows an example of the configuration of a waste tracing system 10. Fig. 1A shows an example of tracing and simulating GHG emissions.

[0014] In Figure 1A, a waste tracing system 10 (hereinafter sometimes simply referred to as the "system") has an integrated server 11, a partial model server 12, a tracer (tracing means) 13, a simulator (simulation means) 14, and an execution environment allocation means 16. For convenience of explanation, the integrated server 11 and the partial model server 12 may be collectively referred to as the "overall model." For simplicity, Figure 1A shows three partial model servers 12a, 12b, and 12c, but the number of partial model servers 12 is not limited to three. The overall model estimates or calculates GHG emissions.

[0015] The estimation or calculation is realized by various methods such as mathematical formulas, simulation languages, rules of thumb, and past case databases. Here, estimation and calculation are collectively referred to as "calculation." Furthermore, the method of expressing the calculation is referred to as "calculation model." The integrated server 11 has a calculation model that calculates the relationship between raw materials, fuels, and chemical substances input or emitted by each business in the real world. The partial model server 12 has a calculation model that calculates the GHG emissions of individual businesses or individual equipment. An example configuration of the partial model server 12 is shown later in Figure 1B.

[0016] The tracer 13 tally up the actual GHG emissions calculated by the integrated server 11 and the partial model server 12 for each business. The operator 15 operates the simulator 14 and receives simulation results from the simulator 14. The simulator 14 calculates the GHG emissions predicted when the actual physical quantity time-series input signals and parameters change using the calculation means provided by the integrated server 11 and the partial model server 12, and provides the calculated amounts to the operator 15.

[0017] Here, the physical quantity time-series input signal refers to a signal indicating a time-dependent physical quantity, such as the amount of raw material input, the amount of fuel consumed, the amount of energy provided to a heater, or the power used by a utility within the plant. A parameter refers to a quantity treated as a constant, such as combustion efficiency or reaction volume. When a time-dependent physical quantity cannot be directly measured by a sensor or the like and is assumed to be a constant value, the physical quantity assumed to be a constant value is treated as a parameter. For example, when performing a simple calculation assuming the excess air ratio in a combustor to be a constant value, the excess air ratio is treated as a parameter.

[0018] 1A shows one tracer 13 and one simulator 14. However, it is also possible to connect a plurality of each of these. The execution environment allocation means 16 An execution environment for tracing is assigned to the tracer 13, and an identification number for identifying the tracer 13 is issued, and an execution environment for simulation is assigned to the simulator 14, and an identification number for identifying the simulator 14 is issued. Identification numbers are associated with all physical quantity time-series input signals and parameters that the tracer 13 or simulator 14 transmits to the integrated server 11, as well as all outputs received from the integrated server 11, making them distinguishable.

[0019] The execution environment management table 17 stores identification numbers assigned to the tracer 13 and the simulator 14. The end-user business 18 is a business that ultimately uses the GHG emissions tallied by the tracer 13, such as a certification body that certifies carbon footprints or a market that trades CO2 emissions rights.

[0020] In Fig. 1A, solid arrows represent the flow of general physical quantity time-series input signals. Dashed arrows represent identification numbers and signals requesting identification numbers from the execution environment allocation means 16. Arrows with diagonal lines represent multiple signals collectively. Arrows with "(P)" represent parameters.

[0021] The sensor 101 is a fuel flow meter, concentration sensor, or the like that acquires obtainable physical quantities in the real world as information such as electrical signals. The edge device 102 digitizes the information acquired by the sensor 101 and transmits it to the tracer 13 via a wide area communication network (not shown in FIG. 1A). The sensor / model correspondence table 104 describes the correspondence between sensor information and calculation models.

[0022] The parameter setting means 105 receives the identification number and parameters from the simulator 14. During the tracing operation, the parameters are set to actual values ​​and are rarely changed, so transmission of the parameters from the simulator 14 to the parameter setting means 105 can be omitted. In this case, the parameter setting means 105 sets the parameters of the overall model by itself.

[0023] The physical quantity time-series input signal setting means (input setting means) 106 receives the identification number and the physical quantity time-series input signal from the tracer 13 or the simulator 14. The physical quantity time-series input signal setting means 106 sets the received information as an input for the overall model.

[0024] The overall model template 107 is a calculation model owned by the integrated server 11, and describes the input / output relationship between the overall model template 107 and the partial model servers 12. The overall model template 107 delegates part of the calculation to the partial model servers 12. The virtual partial models 109a, 109b, and 109c are parts of the overall model template 107 whose calculations are delegated to the partial model servers 12. The partial model server table 108 describes the correspondence between the virtual partial models 109 and the partial model servers 12 to which the calculations are delegated.

[0025] 1A correspond to the partial model servers 12a, 12b, and 12c, respectively. Note that the number of partial model servers 12 does not necessarily have to match the number of virtual partial models 109, and one partial model server 12 may be configured to handle two or more virtual partial models 109.

[0026] The previous value saving means 110 is provided when the overall model template 107 has a feedback element. The previous value saving means 110 temporarily stores the output to be fed back and outputs it to the physical quantity time-series input signal setting means 106 when the calculation model is executed next time. The physical quantity time-series input signal setting means 106 outputs the output obtained when the calculation model was executed last time to the overall model template 107 as a physical quantity time-series input signal.

[0027] Such feedback is necessary, for example, when a thermal power generator burns a mixture of fossil fuel and biofuel. The thermal power generator generates a predetermined amount of power in accordance with market demand. The ratio of the amount of fossil fuel to the amount of biofuel can be freely set under the constraint that the total amount of power generated equals the predetermined amount of power. In this case, a method is used to calculate the ratio of the amount of fossil fuel to the amount of biofuel by convergence calculation so that it is consistent with the amount of biofuel produced by the biofuel producer.

[0028] The part model interfaces 111a, 111b, and 111c exchange information between the virtual part model 109 and the part model server 12. The three part model interfaces 111a, 111b, and 111c in Fig. 1A correspond to the part model servers 12a, 12b, and 12c, and the virtual part models 109a, 109b, and 109c. In general, the number of part model servers 12 corresponds to the number of part model interfaces 111.

[0029] The partial model interface 111 receives a physical force time-series input signal from the virtual partial model 109, and receives parameters and an identification number from the parameter setting means 105. Thereafter, the partial model interface 111 transmits the received information to the partial model server 12. Thereafter, the partial model interface 111 receives the calculation result and identification number of the partial model server from the partial model server 12. Thereafter, the partial model interface 111 outputs the received calculation result of the partial model server 12 to the virtual partial model 109. In the whole model template 107, the virtual partial model 109 appears to have the same input and output as a calculation model (partial model 121 described later) existing in the partial model server 12.

[0030] The output means 112 transmits the integrated results of the calculations made by the overall model template 107 and the partial model server 12, i.e., the calculation results of the overall model, to the tracer 13 or simulator 14 corresponding to the identification number. In other words, a calculation result corresponding to the settings of the tracer 13 is returned in response to a request from the tracer 13, and a calculation result corresponding to the settings of the simulator 14 is returned in response to a request from the simulator 14. When the overall model template 107 has a feedback element, the output means 112 also outputs the calculation result to the previous value saving means 110.

[0031] The execution process monitoring means 113 monitors the execution state of the partial model server 12 obtained from the partial model interface 111. If it is expected that the system load will exceed a predetermined value, the execution process monitoring means 113 notifies the simulator 14 of information such as the system load. The operation of the execution process monitoring means 113 will be described later.

[0032] 1B shows an example of the configuration of the partial model server 12. The partial model 121 is a calculation model corresponding to the virtual partial model 109. The partial model management means 122 creates a physical quantity time-series input signal to the partial model 121, and collectively transmits the outputs from the partial model 121 to the integrated server 11. The parameter database operation means 123 creates and stores parameters to be used in the calculation, and transmits them to the simulator 14.

[0033] When there are missing values ​​in the physical quantity time-series input signals, the missing input complementing means 124 creates a tentative physical quantity time-series input signal using a valid physical quantity time-series input signal obtained in the past. The model validity monitoring means 125 compares the calculation results of the partial model 121 with actual values ​​and monitors the validity of the calculation results. The actual values ​​can be obtained, for example, from actual measurement information of the sensor 101 transmitted from the integration server 11 to the partial model server 12, or from information obtained by processing the actual measurement information. The operation of the model validity monitoring means 125 will be described later.

[0034] The partial model management means 122 is composed of an operation selection means 131, a physical quantity time-series input signal creation means 132, and an input / output correspondence means 133. The operation selection means 131 refers to the identification number, parameters, and physical quantity time-series input signal sent from the partial model interface 111, and if a change in a parameter is requested, it sends the value of the parameter to be changed to the parameter database operation means 123.

[0035] There are cases where the physical quantity time-series input signals input to the operation selection means 131 satisfy all of the physical quantity time-series input signals required by the partial model 121. In this case, the physical quantity time-series input signal creation means 132 transmits the physical quantity time-series input signals input to the operation selection means 131 as they are to the partial model 121. There are also cases where the physical quantity time-series input signals input to the operation selection means 131 have missing values ​​compared to the physical quantity time-series input signals required by the partial model 121. In this case, the physical quantity time-series input signal creation means 132 acquires implicit input values ​​from the missing input complementation means 124, complements the missing physical quantity time-series input signals, and outputs the complemented physical quantity time-series input signals to the partial model 121. The input / output correspondence means 133 associates the calculation result output of the partial model 121 corresponding to the physical quantity time-series input signals with an identification number and transmits the result to the integration server 11.

[0036] The parameter database operation means 123 is composed of a parameter database 134, calculation parameter setting means 135, and parameter notification means 136. The parameter database 134 stores parameters used in calculations by the partial model 121 in association with an identification number, and outputs parameters corresponding to the identification number to the calculation parameter setting means 135. The calculation parameter setting means 135 sets calculation parameters for the partial model 121 for each identification number, thereby running a simulation in which the calculation parameters of the partial model 121 are changed in parallel with tracing. In this way, the common partial model 121 can simultaneously satisfy the different requirements of tracing and simulation.

[0037] The parameter notification means 136 notifies the simulator 14 of parameter information held by the partial model server 12. Here, the simulator 14 receives the meanings and setting values ​​of the parameters that can be set as parameter information, and the operator 15 knows the parameters that can be changed. Furthermore, if the owner of the partial model server 12 has parameters that he does not want to disclose as trade secrets, hidden know-how, etc., the parameter notification means 136 can easily prohibit them from being disclosed to the outside.

[0038] Such a data hiding mechanism is essential when configuring a system in cooperation with multiple industries, as in this embodiment. Furthermore, by providing such a mechanism, it becomes easier to provide the calculation environment using the partial model server 12 to the outside. Such a mechanism can be easily realized by separating the integrated server 11 and the partial model server 12 and configuring a waste tracing system as a whole.

[0039] The missing input complementing means 124 is composed of an implicit input creating means 137 and an implicit input database 138. When the waste tracing system 10 is performing a tracing operation and the physical quantity time series input signals acquired by the sensors 101 satisfy all of the physical quantity time series input signals required by the partial model 121, the implicit input creating means 137 acquires the physical quantity time series input signals acquired by the sensors 101 as they are. The acquired physical quantity time series input signals are processed and stored in the implicit input database 138.

[0040] There are various methods for processing the physical quantity time-series input signal. As the physical quantity time-series input signal, a value that is suitable as a value that represents normal operation is selected. For example, the average value when the rate of change of the physical quantity time-series input signal is equal to or less than a certain value, the mode value of the physical quantity time-series input signal at a preset time, etc. are suitable examples.

[0041] When there is a missing value in the physical quantity time-series input signal, the implicit input database 138 outputs the processed physical quantity time-series input signal to the physical quantity time-series input signal creating means 132. As is clear from the above, the implicit input creating means 137 calculates an implicit input value from the history of the physical quantity time-series input values ​​from the tracer 13. The physical quantity time-series input signal creating means 132 uses the implicit input value calculated by the implicit input creating means 137 in place of the missing part of the input from the simulator 14. "Implicit" means that the missing part is filled in without the knowledge of the operator.

[0042] The functions of the waste tracing system that are primarily performed by software have been explained above. The relationship between software and hardware will be described later. Next, the input / output relationship between the overall model template 107 and the partial model 121 will be explained using a simple example.

[0043] Figure 2 is an example of the overall model template 107. Figure 2 shows an example in which a system consisting of a combustor, a water electrolysis device, and a methanation device traces GHG emissions. As specific examples of the virtual partial model 109, a combustor virtual partial model 109a, a water electrolysis device virtual partial model 109b, and a methanation device virtual partial model 109c correspond to the combustor, the water electrolysis device, and the methanation device, respectively, in the real world.

[0044] In Figure 2, CH4(f) is methane fuel derived from fossil fuels. CH4(r) is methane fuel derived from renewable energy (recycled). CH4(e) is unburned methane released into the atmosphere. CO2(r) is carbon dioxide that is captured (recycled) and sent to the methanation unit. CO2(e) is carbon dioxide that is not captured and released into the atmosphere. H2O is water input to the water electrolysis unit. H2 is hydrogen generated by the electrolysis unit's electrolysis of water.

[0045] The combustor takes CH4(f) and CH4(r) as input and generates CO2 through a combustion reaction. The remaining residue is CH4(e). Some of the generated CO2 is recovered by a CO2 recovery unit and becomes CO2(r), while the CO2 that cannot be recovered becomes CO2(e). The methanation unit uses H2 and CO2(r) generated by the water electrolysis unit as raw materials and generates CH4(r) through the Sabatier reaction, etc. CH4(r) is sent back to the combustor and reused as fuel. For simplicity, the electricity used by the water electrolysis unit and the energy required by the methanation unit to maintain a reaction environment are omitted.

[0046] In this example, the amount of CO2(r) generated by the combustor at a past time affects the amount of CH4(r) generated by the methanation unit at the present time, and the current amount of CH4(r) becomes the CHR(r) input to the combustor in the future, so the previous value storage means 110 is essential. In this example, an appropriate initial value for CH4(r) is set at the start of the calculation, and the final amounts of CH4(f) and CH4(r) are determined by performing a convergent calculation.

[0047] FIG. 3 is a diagram showing the input / output relationship between the overall model template 107 and the partial model 121 when focusing on the combustor. For simplicity, only the combustor portion is shown. The initial values ​​of CH4(f) and CH4(r) obtained from the physical quantity time-series input signal setting means 106 are sent to the combustor virtual partial model 109a. The combustor virtual partial model 109a assigns the name of the partial model server 12 used in the calculation and variable name information in the partial model 121a according to the contents of the sensor / model correspondence table 104. As a result, the input CH4(f) and CH4(r) are associated with the combustor partial model 121a.

[0048] 1A interprets the name of the allocated partial model server 12 and the variable name information in the partial model 121a. Then, the partial model interface 111a transmits the physical quantity time-series input signals (here, CH4(f) and CH4(r)) in the partial model 121a of the combustor to the correct delegated partial model 121a of the combustor via the wide area communication network 141.

[0049] The combustion calculation model 142 calculates the amounts of CO2 and CH4 (molar amounts) using parameters obtained from the parameter database 134a. Generally, GHG gases are managed in terms of the CO2 converted mass emitted per unit time. Therefore, the combustion calculation model 142 converts the amounts of CO2 into mass. A CO2 mass flow rate conversion means 143 and a CH4 mass flow rate conversion means 144 calculate the mass flow rate of CO2 (CO2_kgph: mass of CO2 generated per hour) and the mass flow rate of CH4 (CH4_kgph: mass of CH4 generated per hour), respectively. The partial model interface 111a shown in FIG. 1A associates the calculated CO2_kgph and CH4_kgph with the output of the combustor virtual partial model 109a.

[0050] The reserve unit 145 of the overall model template 107 performs the calculation within the integrated server 11 without delegating the calculation to the partial model server 12. In this example, the reserve unit 145 calculates CO2(r) and CO2(e) based on the calculated CO2_kgph and the recovery rate set as a parameter of the CO2 recovery device. No subsequent processing by the reserve unit 145 is set for CH4_kgph. Therefore, the calculated CH4_kgph becomes CH4(e) as is. CO2(r) is sent to the subsequent methanation virtual partial model 109c, where calculation is performed in the partial model 121c (not shown).

[0051] When the storage unit 145 is present as in this example, the integrated server 11 includes a parameter database operation means 123 shown in Fig. 1B in order to set the parameters of the storage unit 145. However, for simplicity, the description of this configuration is omitted here.

[0052] In this example, the GHG gases to be traced and simulated are CO2(e) and CH4(e) generated by the combustor, and CO2(e) and CH4(e) generated by the methanation unit. These values ​​are output to the output means 112 in Figure 1A and then sent to the tracer 13 and simulator 14. The tracer 13 and simulator 14 take into account the global warming potential and then tally up the final GHG amounts for each business operator.

[0053] The tracing operation according to this embodiment will now be described. FIG. 4A shows the data flow between the components in the first half of the trace operation. FIG. 4B shows the data flow between the components in the second half of the trace operation. FIG. 5A is a flowchart of the first half of the tracing operation. FIG. 5B is a flowchart of the second half of the tracing operation.

[0054] (First half of the trace operation) The first half of the tracing operation will be explained with reference to Figures 4A and 5A. In the explanation of Figure 4A, reference will be made to Figure 6. "SXXX" in Figure 4A corresponds to "Step SXXX" in Figure 5A. Note that steps S206 and S207 are operations performed within integrated server 11, and are omitted from Figure 4A.

[0055] In FIG. 4A, sensor 101a measures fuel flow rate. Sensor 101b measures CO2 concentration. Sensor 101c measures CH4 concentration. Sensor 101a is an intelligent sensor that can transmit information directly to tracer 13. Sensors 101b and 101c are ordinary sensors that can transmit information to tracer 13 via edge device 102. Each sensor is assigned an address that can be identified externally. The correspondence between each variable in overall model template 107 and partial model 121 and the measurement value by sensor 101 is written in sensor / model correspondence table 104 in tracer 13.

[0056] FIG. 6 is a diagram showing an example of the sensor / model correspondence table 104. FIG. 6 shows an example of the correspondence between a sensor address, a partial model server ID that uniquely identifies a partial model server 12, and an input variable name in a partial model. There are various methods for specifying a sensor address, but MQTT (Message Queuing Telemetry Transport) is assumed here. It is also possible to specify not only the input variable name in the partial model 121, but also a variable in the overall model template 107 as necessary. In this case, for example, an ID that identifies the integrated server, such as "0", can be assigned to the partial model server ID.

[0057] The sensor / model correspondence table 104 has a column showing the type of process. Specific examples of the type of process are classifications by industry, such as electrolysis and combustion. This allows the tracer 13 to tally GHG by industry. In addition, classification columns for business names, etc., may be provided as needed.

[0058] 5A, a trace operation is started in response to the occurrence of a trace execution event, such as polling at regular intervals or an interrupt signal generated when there is a sensor input. In step S202, the tracer 13 acquires a sensor value from the sensor 101 having the sensor address listed in the sensor / model correspondence table 104.

[0059] In step S203, the tracer 13 requests the execution environment allocation means 16 to issue an identification number. The execution environment allocation means 16 refers to the execution environment management table 17 and issues a new identification number (here, "t6953") that does not overlap with any existing identification numbers. The execution environment management table 17 stores the requesting client name and the type of processing in association with the identification number. The client name here is a unique MAC (Media Access Control) address. The type of processing is "tracer."

[0060] In step S204, the execution environment allocation means 16 assigns the issued identification number to the tracer 13. Thereafter, the physical quantity time-series input signal and the calculation result are given an identification number and are exchanged between the integration server 11 and the partial model server 12. For simplicity, the physical quantity time-series input signal will be simply referred to as "input" or "input value" unless misunderstanding occurs.

[0061] In step S205, the tracer 13 transmits the issued identification number and the sensor value obtained from the sensor 101 to the integrated server 11. At this time, the tracer 13 also transmits the sensor address to the integrated server 11. In step S206, the integration server 11 refers to the sensor / model correspondence table 104 to obtain the partial model server ID and the input variable name. In step S207, the integrated server 11 refers to the partial model server table 108 and associates the virtual partial model 109 with the partial model server 12. Here, it is assumed that the virtual partial model 109 is associated with the partial model server 12a having the identifier "ID:232".

[0062] In step S208, the integrated server 11 transmits the identification numbers and input values ​​to the partial model server 12 in accordance with the execution order of the calculations described in the overall model template 107. Note that the input values ​​here are associated with the input variable names in the partial model of Fig. 6, and in the partial model 121a, are set to the values ​​of the variables defined in the partial model 121a. In step S209, the partial model server 12a receives the identification number and the input value from the integration server 11.

[0063] In step S210, the partial model server 12a sets parameters from the parameter database 134a and outputs the parameters to the partial model 121a. In step S211, the partial model server 12a obtains the calculation results of the partial model 121a using the input values ​​received in step S209.

[0064] In step S212, the partial model server 12a processes the input values ​​and updates the implicit input database 138a. Since actual sensor values ​​are input during the tracing operation, as described above, it is possible to obtain a representative input, for example, an average value when the rate of change of the physical quantity time-series input signal becomes equal to or less than a certain value, or a mode value of the physical quantity time-series input signal at a preset time. The information in the implicit input database 138a updated here is used during the simulation described below. In step S213, the partial model server 12a transmits the calculation result of the partial model 121a to the integration server 11 together with the identification number.

[0065] (Later part of the trace operation) The latter half of the tracing operation will be explained with reference to Figures 4B and 5B. "SXXX" in Figure 4B corresponds to "Step SXXX" in Figure 5B. For ease of understanding, Figure 4B also shows "S213" in the first half of the tracing. On the other hand, steps S214, S215, and S218 are processes performed within the integrated server 11 or the tracer 13, and therefore are omitted from Figure 4B.

[0066] 5B, the integrated server 11 refers to the entire model template 107 and determines whether there is a subsequent process (virtual partial model 109) that uses the calculation result received from the partial model server 12a as input. If there is a subsequent process ("Yes" in step S214), the process proceeds to step S215. Otherwise ("No" in step S214), the process proceeds to step S217.

[0067] In step S215, the integration server 11 refers to the partial model server table 108 and associates the subsequent virtual partial model 109 with the subsequent partial model server 12b. The identification number of the subsequent partial model server 12b here is "ID:781".

[0068] In step S216, the integration server 11 transmits the calculation result and the identification number received in step S213 to the subsequent partial model server 12b. If the overall model template 107 includes a reserve unit 145, the integration server 11 extracts an execution order from, for example, the dependency relationship between the virtual partial model 109 in the overall model template 107 and the reserve unit 145, and adds a procedure for performing calculations in accordance with that order. Then, the process returns to step S209.

[0069] For simplicity of explanation, the flowchart for the case where the reservation unit 145 is included is omitted here. The processing of steps S209 to S216 is repeated until all outputs of the overall model template 107 are fulfilled. Now, assume that all outputs of the overall model template 107 have been fulfilled.

[0070] In step S217, the integrated server 11 transmits the calculation results of the overall model template 107 to the tracer 13 that transmitted the identification number "t6953." The calculation results transmitted here are the targets of tracing, and are, for example, the hourly emissions of CO2 and CH4 calculated by each partial model server 12. To identify the industry from which the values ​​were emitted, the partial model server ID of the calculation source is added to each output of step S217.

[0071] In step S218, the tracer 13 compiles output values ​​with the same identification number (here, "t6953") and tallies the GHG emissions from each business. Generally, GHG emissions are expressed as the mass of carbon dioxide emitted per hour (t-CO2 / h or kgCO2 / h). Here, since CO2 and CH4 are the subject of GHG, the tracer 13 performs a process of multiplying the CH4 emission mass by a predetermined global warming potential for each business and adding the result to the CO2 emission mass.

[0072] In step S219, the tracer 13 transmits the GHG emission amount information to the end-use business operator 18. At this time, the tracer 13 may display the GHG emission amount information on an arbitrary device. In step S220, the tracer 13 sends an end signal to the execution environment allocation means 16. The execution environment allocation means 16 then clears the identification number "t6953" and its related information, thereby ending the tracing operation.

[0073] Next, the simulation operation according to this embodiment will be described. FIG. 7A shows the data flow between the components in the first half of the simulation operation. FIG. 7B shows the data flow between the components in the latter half of the simulation operation. FIG. 8A is a flowchart of the first half of the simulation operation. FIG. 8B is a flowchart of the second half of the simulation operation. Detailed explanations of processes common to the tracing operation and the simulation operation will be omitted.

[0074] (First half of the simulation) The first half of the simulation operation will be explained with reference to Fig. 7A and Fig. 8A. In the explanation of Fig. 8A, Fig. 9 will be referred to. "Sxxx" in Fig. 7A corresponds to "Step Sxxx" in Fig. 8A. Note that step S231 is a process inside the simulator 14a. Step S234 is an operation by the operator 15a. Step S240 is an operation inside the integration server 11. Steps S243 and S245 are processes inside the partial model server 12. Therefore, these descriptions are omitted in Fig. 7A.

[0075] The simulators in Fig. 7A are the simulator 14a and the simulator 14b. Fig. 7A illustrates a case where the simulator 14a newly executes a simulation in a state where both the tracer 13 and the simulator 14b are operating.

[0076] In step S231 of FIG. 8A, a simulation execution event is initiated in response to a simulation start signal or the like from the simulator 14a, thereby starting the simulation operation. In step S232, the simulator 14a receives information about the overall model template 107 from the integration server 11. This process enables the simulator 14a to set input variables and parameters of the overall model template 107 according to the intentions of the operator 15a.

[0077] In step S233, the simulator 14a receives information about available partial model servers 12 from the integrated server 11. Here, "available" means that the partial model servers 12 permit the parameter database 134 to be made publicly available, and that no events such as a server down have occurred. The partial model server table 108 in the integrated server 11 stores information about available partial model servers 12, and the simulator 14a obtains information about available partial model servers 12.

[0078] Fig. 9 is a diagram showing an example of the partial model server table 108. Fig. 9 stores simulation supportability information for each partial model server ID. For example, by setting "FALSE" as the simulation supportability information, it can be seen that the corresponding partial model server 12 has refused to make the model public.

[0079] FIG. 9 also stores abnormality detection information related to the partial model server ID. The abnormality detection information is the alive status of the partial model server 12, obtained from the partial model server 12 at regular intervals, for example. "TRUE" indicates that an abnormality has occurred. In this case, some kind of abnormality has occurred in the partial model server with the partial model server ID "233", and new simulations cannot be executed using this partial model server 12. Note that alive status monitoring can be easily achieved using a network monitoring tool, etc.

[0080] By following the procedure up to this point, the operator 15a can obtain information on the partial model server 12 that can be used for simulation. In step S234 of FIG. 8A, the operator 15a selects the partial model servers 12a and 12b to be used in the simulation. In step S235, the operator 15a refers to the parameter databases 134a and 134b (of which the parameter database 134b is not shown) in the partial model servers 12a and 12b, and sets parameters to be used in the simulation.

[0081] In step S236, the operator 15a similarly sets input values ​​for the simulator 14a. In step S237, the simulator 14a requests the execution environment allocation means 16 to issue an identification number. This process corresponds to the process of S203 in the tracing operation. Here, it is assumed that the tracer (identification number t6953) and simulator (identification number s17) are operating, so an identification number "s18" different from those is issued.

[0082] In step S238, the execution environment allocation means 16 transmits the identification number "s18" to the simulator 14a. In step S239, the simulator 14a transmits the identification number, the parameters, and the input values ​​to the integrated server 11.

[0083] In step S240, the integration server 11 refers to the partial model server table 108 to associate the virtual partial model 109 with the partial model server 12. In step S241, the integration server 11 transmits the identification number, parameters, and input values ​​to the partial model server 12a in accordance with the execution order of the whole model template 107. In step S242, the partial model server 12a obtains the identification number, parameters, and input values ​​from the integration server 11.

[0084] Steps S240, S241, and S242 in the simulation operation correspond to steps S205, S207, and S208 in the tracing operation, respectively. However, there are two main differences between the tracing operation and the simulation operation: First, in the tracing operation, input from the sensor 101 becomes input to the overall model template 107, and therefore a sensor / model correspondence table 104 that describes the correspondence between the sensor 101 and the overall model template 107 is referenced. In contrast, in the simulation operation, the operator 15 directly sets input variables while looking at the information in the overall model template 107. Second, in the tracing operation, parameters are fixed, whereas in the simulator operation, parameter change processing is performed.

[0085] In step S243, the partial model server 12a determines whether all the parameters of the partial model 121a have been set. If all the parameters of the partial model 121a have been set (step S243 "Yes"), the process proceeds to step S245. Otherwise (step S243 "No"), the process proceeds to step S244.

[0086] In step S244, the partial model server 12a receives the missing parameters from the parameter database 134a. In step S245, the partial model server 12a determines whether all input values ​​of the partial model 121a have been set. If all input values ​​of the partial model 121a have been set (step S245 "Yes"), the process proceeds to step S248. Otherwise (step S245 "No"), the process proceeds to step S246.

[0087] In step S246, the partial model server 12a compensates for the implicit input value from the implicit input database 138a and transmits it to the integration server 11. In step S247, the integration server 11 fills in the missing parts of the input values ​​with implicit input values ​​to make them complete input values, and then transmits them to the partial model server 112a.

[0088] (Later part of the simulation) The latter half of the simulation operation will be explained with reference to Figures 7B and 8B. "SXXX" in Figure 7B corresponds to "Step SXXX" in Figure 8B. Note that step S248 is a process performed within the partial model server 12a. Steps S250 and S251 are processes performed within the integration server 11. Step S254 is a process performed within the simulator 14a. Therefore, these steps are omitted from Figure 7B.

[0089] In step S248 of FIG. 8B, the partial model server 12a calculates the output values ​​using the input values. In step S249, the partial model server 12a transmits the identification number and the calculation result to the integration server 11. Unlike the tracing operation, in the simulation operation, the input values ​​from the simulator 14a are not necessarily based on reality, and therefore, the implicit input database is not updated as in step S212 of Fig. 5A.

[0090] In step S250, the integrated server 11 determines whether there is a subsequent process that uses the calculation result received from the partial model server 12a as input. If there is a subsequent process (step S250 "Yes"), the process proceeds to step S251; otherwise (step S250 "No"), the process proceeds to step S253. In step S251, the integration server 11 refers to the partial model server table 108 to associate the subsequent virtual partial model 109b with the subsequent partial model server 12b. In step S252, the integrated server 11 transmits the received calculation result and identification number to the subsequent partial model server 12b. Then, the process returns to step S242. Steps S242 to S252 are repeated until all outputs of the overall model template 107 are satisfied. Now, assume that all outputs of the overall model template 107 have been satisfied.

[0091] In step S253, the integrated server 11 transmits the calculation results of the overall model template 107 to the simulator 14a that transmitted the identification number "s18." In step S254, the simulator 14a collects output values ​​having the same identification number (here, "s18") and totals the GHG emissions. In step S255, the simulator 14a displays the GHG emission amount information to the operator 15 (or to any device). In step S256, the simulator 14a transmits an end signal to the execution environment allocation means 16. The execution environment allocation means 16 then deletes the identification number "s18" and its related information, thereby ending the simulation operation.

[0092] The configuration and operation of a waste tracing system according to this embodiment have been described above, taking GHG tracing as an example. In this embodiment, in a distributed server configuration made up of an integrated server 11 and a partial model server 12, a parameter database 134 is placed in the partial model server 12, and parameters and physical quantity time series inputs are switched depending on the calculation environment.

[0093] Generally, power companies, energy regeneration companies, and other businesses that actually convert energy and are involved in the generation and absorption of GHGs have detailed calculation models and know-how for each facility. Since detailed calculation models and know-how are the strengths of each business, there is little motivation for each business to disclose them to the outside. Therefore, it is desirable for each business to have a partial model server 12 and to set up a parameter database 134 within the partial model server 12.

[0094] As a result, in order to realize a waste tracing system, each business operator does not need to provide all of the calculation models etc. to the business operator that provides the GHG emission calculation service, but only needs to provide the calculation service. In this configuration, as shown in Figures 1A and 1B, by having a means for appropriately switching parameters and physical quantity time-series input according to the identification number, it becomes possible to perform tracing while simultaneously running a simulation using the same partial model 121. This reduces the introduction cost compared to when separate models are prepared for the tracer and simulator.

[0095] 1A and 1B show an example of a configuration in which the integrated server 11, partial model server 12, tracer 13, simulator 14, and execution environment allocation means 16 are all realized by separate hardware. However, if a configuration exists in which a partial model server 12 separate from the integrated server 11, other embodiments can be adopted without changing the spirit of the present invention.

[0096] Fig. 10 is a diagram showing a modified example of the waste tracing system 10. Here, the integrated server 11 also provides the functions of the tracer 13 and the execution environment allocation means 16. That is, in Fig. 10, the integrated server 11, tracer 13, and execution environment allocation means 16 are configured in a single housing. Also, some of the calculations performed by the partial model server 12 in Fig. 1A are performed by the holding unit 145 of the integrated server 11. In this case, the integrated server 11 performs some of the parameter setting and implicit input value creation functions performed by the partial model server 12 in Fig. 1A.

[0097] For this reason, the integrated server 11 has a parameter database 134 and an implicit input database 138, as well as management means for them (parameter database operation means 123, missing input completion means 124, implicit input creation means 137, etc.: not shown). This configuration reduces the amount of hardware required and reduces implementation costs when the amount of waste being handled is low and the update frequency is low.

[0098] 11 is a diagram illustrating the operation of the execution process monitoring means 113. It is assumed that a trace operation is started by a trace execution trigger 152 in a trace execution period 151. It is assumed that a total of n tasks are executed within the trace execution period 151. Here, a task means either a trace operation or a simulation operation. It is assumed that on average, "n-1" simulation operations are executed in response to one trace operation.

[0099] Since the same overall model template 107 and the same partial model 121 are used for calculations in the tracing operation and the simulation operation, it can be simplified that both the tracing operation and the simulation operation occupy the system for the same processing time 153. In cycle 1, tracing operates. In cycles 2 to 4, in addition to tracing, one to three simulations operate on average. In other words, tracing is always operating.

[0100] The system occupation time from when the trace operation is started by the trace execution trigger 152 until the trace operation or the final simulation operation ends is designated as 154. The system occupation times for cycle 1, cycle 2, cycle 3, and cycle 4 are designated as 154a, 154b, 154c, and 154d, respectively. In cycles 1, 2, and 3 in Fig. 11, where the average number of simulation executions is 2 or less, the system occupation time 154 is shorter than the trace execution cycle 151, and therefore the system operates without failure.

[0101] On the other hand, in cycle 4, the system occupation time 154d is longer than the trace execution cycle 151. As a result, part of simulation 3 remains unprocessed and is delayed until the next time it can be executed. Reference numeral 155 denotes the delayed processing time. If the number of tasks "n" running during the next trace execution cycle 151 is 2 or less, the delayed tasks can be executed within the trace execution cycle 151. However, if "n" is consistently 3 or more, the delayed processing time 155 accumulates and the requested operation will not be executed.

[0102] The execution process monitoring means 113 addresses this issue. In its simplest form, the execution process monitoring means 113 has information on the processing time 153 "τ" and constantly monitors the number of tasks "n" received by the integrated server 11. When the relationship "(n+1)τ>T" holds between the execution time 153 and the trace execution period 151 "T," the execution process monitoring means 113 notifies the simulator 14 that "the load is high and calculation may not be possible." The relationship "(n+1)τ>T" means that if one task is added to the number of currently executing tasks "n," the execution time of the "n+1" tasks will be longer than the trace execution period 151. In other words, the execution process monitoring means 113 constantly monitors the ratio of the trace execution time "(n+1)τ" to the trace execution period 151 "T." If the ratio is expected to exceed 100%, the execution process monitoring means 113 notifies the simulator 14, suppresses the introduction of new simulations, and prevents excessive system load.

[0103] Furthermore, when there are multiple traces, each with the highest priority and equal processing time, techniques such as rate monotonic scheduling can also be applied. This calculates the load factor limit as "n × (pow(2,1 / n)-1)." For example, when n → ∞, a load factor of 69% is the limit value at which the system will not fail. For example, the execution process monitoring means 113 monitors the ratio of system occupancy time 154 to the trace execution period 151. If this ratio exceeds 69%, the execution process monitoring means 113 can also issue a warning to the simulator 14 that "the system may not be able to meet the simulation requirements." Note that pow(x,y) is a function representing x raised to the power y.

[0104] By providing the execution process monitoring means 113, if the system load is heavy, it is possible to encourage the user to reduce the simulation load, which helps to improve the stability of the system. Furthermore, if the load rate is constantly high, it is possible to consider measures such as updating the system, which helps to improve the availability of the system.

[0105] FIG. 12 is a diagram illustrating the operation of the model validity monitoring means 125. The horizontal axis in FIG. 12 represents time. The vertical axis in FIG. 12 represents the calculation results or measured values ​​in the overall model template 107 and the partial model 121, specifically, the theoretical calculation results of the CO2 emissions, the measured CO2 emissions, etc. The model validity monitoring means 125 monitors the values ​​on the vertical axis in chronological order and creates the graph in FIG. 12. When there is no sensor for measuring CO2 emissions, the model validity monitoring means 125 monitors the CO2 concentration sensor and the exhaust gas flow sensor, and treats the value estimated from these as the measured CO2 emissions.

[0106] 12, ◯ indicates a calculation result, × indicates an actual measurement value, and error bars indicate the range of calculation results under the same input conditions. △ indicates a moving average of the actual measurement values ​​(a moving average for a predetermined number of times that is set appropriately). The model validity monitoring means 125 acquires or calculates these values ​​in synchronization with the trace execution trigger 152.

[0107] Since the tracing operation is constantly performed, the partial model server 12 can acquire the time history of past actual measurement values ​​and calculation results. Therefore, the model validity monitoring means 125 can create a frequency distribution of past calculation results for each input condition. The range (error bar) of past calculation results is, for example, the past calculation result ±3σ (σ is the standard deviation of the actual measurement value). The model validity monitoring means 125 determines that some abnormality has occurred when the actual measurement value exceeds the error bar range of the calculation result. Figure 12 shows that when a value 161 outside the error bar range is observed, it suggests that some event has occurred that causes the calculation model to not reflect the actual value, such as a sensor failure or an unexpected change in fuel properties.

[0108] In actual operation, the error bar range deviation value 161 may occur due to the influence of sensor outliers, noise, etc. Therefore, the model validity monitoring means 125 uses a diagnostic criterion that determines an abnormality when the error bar range deviation value 161 is observed a predetermined number of times in succession, when the moving average value (the point represented by a triangle) deviates from the error bar, etc. This makes it possible to quickly grasp the situation when the entire model template 107 or the partial model 121 becomes unrealistic, thereby ensuring the reliability of the system.

[0109] (Relationship between software and hardware) 13 is a diagram showing the hardware configuration of the waste tracing system. Each component included in the waste tracing system 10 of this embodiment (for example, the integrated server 11, partial model server 12, tracer 13, simulator 14, execution environment allocation means 16, etc.) is a general computer 200. The computer 200 has a central control unit 201, an input device 202 such as a keyboard, an output device 203 such as a display, a main memory device 204, an auxiliary memory device 205, and a communication device 206. These are connected by a bus 207.

[0110] Of these, the auxiliary storage device 205 stores various databases, tables, programs, etc. In this embodiment, anything described as "XX means" refers to a program. When the subject is described as "XX means," it means that the central control device 201 reads a program from the auxiliary storage device 205 to the main storage device 204 and executes the processing that is pre-written in that program. The waste tracing system 10 can also realize such software-based processing with hardware.

[0111] (Effects of the waste tracing system of this embodiment)

[0112] Generally, a waste tracing system needs to constantly trace the target waste volume at regular intervals. In order for such a system to also have a simulation function, it is necessary to distinguish whether the system is tracing or simulating. In addition, since it is expected that there will be multiple simulation operators performing the simulation, it is necessary to distinguish between the calculation environment for each simulation operator.

[0113] The waste tracing system of this embodiment has an execution environment allocation means that issues an identification number for each tracer and simulator (i.e., for each service requester), and sets up an independent computing environment for each service requester. The waste tracing system of this embodiment has a parameter database that sets parameters for partial models according to the identification number, and by saving and using different parameters for each identification number, returns to the requester an output corresponding to the input and parameters entered by the service requester. This allows the waste tracing system of this embodiment to mix tracing and multiple simulations.

[0114] In the waste tracing system of this embodiment, the partial model server has a parameter database that sets the parameters of the partial model according to the identification number. As a result, it is possible to conceal some of the parameters corresponding to the know-how of each business from the integrated server provider. In addition, it is possible to prevent the leakage of the calculation know-how of each related business, which has been an obstacle to building a general waste tracing system.

[0115] In the waste tracing system of this embodiment, the integrated server or the partial model server has an implicit input creation means for calculating an implicit input value from the history of input values ​​of the tracer. The physical quantity time-series input signal setting means uses the implicit input value from the implicit input creation means in a portion where no input of the simulator is instructed.

[0116] With this configuration, the waste tracing system of this embodiment uses inputs acquired during tracing to create typical values ​​for inputs under normal circumstances. This allows the operator to set only the inputs that interest them and use the typical values ​​as implicit input values ​​for other inputs, thereby reducing the number of setting items and making the simulation easier to use.

[0117] In the waste tracing system of the present invention, the integrated server has an execution process monitoring means for monitoring the ratio of trace execution time to trace execution cycle. This makes it possible for the waste tracing system of this embodiment to grasp the overall load rate, detect signs of system failure when the load rate exceeds 100%, and take appropriate measures.

[0118] In the waste tracing system according to the present invention, the execution process monitoring means notifies the simulator when the ratio of execution time to the tracing period is expected to exceed 100%, thereby preventing the simulator from submitting tasks that exceed the system's capacity and maintaining the soundness of the system.

[0119] In the waste tracing system of the present invention, the partial model server has a model validity monitoring means for monitoring the validity of the relationship between the actual measurement results and the calculation results. This makes it possible to take measures such as stopping the system or applying an alternative model when the validity of the calculation model can no longer be guaranteed due to a change in the situation, thereby improving the reliability of the system. [Explanation of symbols]

[0120] 10 Waste Tracing System 11 Integrated Server 12, 12a, 12b, 12c Partial Model Server 13 Tracer (Tracing Means) 14 Simulator (simulation means) 16 Execution environment allocation means 105 Parameter setting means 106 Physical quantity time series input signal setting means (input setting means) 107 Whole Model Template 109, 109a, 109b, 109c Virtual Part Model 110 Previous value storage means 111, 111a, 111b, 111c Partial Model Interface 112 Output means 113 Execution process monitoring means 121 Partial Model 125 Model Validity Monitoring Tools 134 Parameter Database 137 Implicit Input Creation Method

Claims

1. A tracing means for tracing the amount of waste for each of the multiple businesses involved; a simulation means for simulating the amount of waste material by varying the amount of raw material input and parameters; an execution environment allocation means for allocating an execution environment for performing the tracing to the tracing means and issuing an identification number for identifying the tracing means, and for allocating an execution environment for performing the simulation to the simulation means and issuing an identification number for identifying the simulation means; a plurality of partial model servers for calculating the amount of waste for each of the businesses based on the amount of raw material input and the parameters; an integration server that integrates the calculation results of the plurality of partial model servers; A waste tracing system comprising:

2. The integrated server a parameter setting means for receiving the identification number and the parameters from the simulation means and transmitting them to the partial model server; an input setting means for receiving the identification number and the raw material input amount from the tracing means and transmitting them to the partial model server; an output means for transmitting the results integrated by the integration server to the tracing means and simulation means identified by the received identification number; a whole model template describing input / output relationships between the plurality of partial model servers; 2. The waste tracing system of claim 1, further comprising:

3. The partial model server a partial model having the raw material input amount as an input and the waste amount as an output; a parameter database that stores parameters used by the partial model in association with the identification number; and The overall model template is a partial model interface for exchanging the raw material input amounts and the parameters between the integrated server and a plurality of the partial model servers; 3. The waste tracing system of claim 2.

4. The partial model server or the integration server an implicit input creation means for calculating an implicit input value from the history of the raw material input amounts; The partial model is Substituting missing raw material input amounts from the tracing means with the implicit input values; 4. The waste tracing system according to claim 3, wherein:

5. The integrated server having an execution process monitoring means for monitoring the ratio of the trace execution time to the trace execution period; 2. The waste tracing system of claim 1.

6. The execution process monitoring means notifying the simulation means that the ratio of the trace execution time to the trace execution period is expected to exceed 100%; 6. The waste tracing system according to claim 5,

7. The partial model server having a model validity monitoring means for monitoring the validity of the relationship between the actual measurement value and the calculation result; 2. The waste tracing system of claim 1.

8. The tracing means and the simulation means Displaying the results calculated by the plurality of partial model servers and integrated by the integration server; 2. The waste tracing system of claim 1.

9. The tracing means, the execution environment allocation means, and the integrated server Constructed in the same housing, 2. The waste tracing system of claim 1.

10. The tracing means of the waste tracing system are: We trace the amount of waste from each of the multiple businesses involved, The waste tracing system simulation means comprises: After varying the raw material input amount and parameters, a simulation of the waste amount is performed. The execution environment allocation means of the waste tracing system assigning an execution environment for performing the tracing to the tracing means and issuing an identification number for identifying the tracing means, and assigning an execution environment for performing the simulation to the simulation means and issuing an identification number for identifying the simulation means; The waste tracing system's partial model server includes: Calculating the amount of waste for each of the businesses based on the amount of raw material input and the parameters; The integrated server of the waste tracing system includes: Integrating the calculation results of the plurality of partial model servers; A waste tracing method comprising:

Images