JP2000184592A - Power system demand scheme generating apparatus and storage medium for recording processing program thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To generate the optimum operation scheme considering the restriction on the power generators. SOLUTION: A partial problem generating means 12 generates a partial subject of each power generator as a partial problem using the Lagrange's alleviating method based on the demand and supply balance of the electrical power. A partial problem combining means 13 combines respective partial problems of power generator group on which the restriction on the power generators is applied. A restriction data memory means 14 stores the restriction data of operation. A restriction condition reflecting means 15 reflects the restriction data on the partial problem and combined partial problems. A means 16 for solving the partial problem and combined partial problem calculates the operation scheme in the minimum cost while satisfying the restriction condition to be considered.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electric power system supply / demand plan creating apparatus for creating an electric power system operation plan from the viewpoint of economical operation of an electric power system, and a recording medium for recording a processing program of the apparatus.

[0002]

2. Description of the Related Art As one of operation plans of an electric power system, there is a problem of starting and stopping a generator. The generator start / stop planning problem is a problem of deciding the operation plan of each generator so that the operation cost is minimized while taking various constraints into consideration, and is very important from the viewpoint of economic operation of the power system. It is a problem.

Conventionally, such an operation plan for starting and stopping a generator has been realized by minimizing a formulated objective function having multiple variables while satisfying the constraints of each variable by using a computer. Mathematical programming is used to minimize the operating costs of

[0004] Methods proposed so far include a method using heuristics, a method using dynamic programming,
There are various methods such as the Lagrangian relaxation method and the branch and bound method. Of these methods, the Lagrangian relaxation method is considered to be the most effective method for the problem of practical scale at present. .

[0005]

However, the conventional method using the Lagrange mitigation method solves a partial problem for each generator by introducing a Lagrange multiplier, and then solves the problem. There are various constraints. However, in the case of the Lagrange mitigation method, the only constraints that can be formulated as the objective function are the constraints related to the generator itself, and all constraints to be considered in the operation of the power system, for example, span between generators. There is a problem that it is difficult to satisfy the constraint conditions and the like.

[0006] For this reason, conventionally, with respect to an operation plan created using the Lagrangian mitigation method, a technician with rich knowledge and experience judges the start and stop of each generator while judging the entire power system including each power plant. The timing is corrected, but the work of this correction and the like requires a lot of trouble and time,
There is a problem that it is not possible to always quickly create an optimal operation schedule without resorting to the judgment of each engineer.

In the Lagrange mitigation method, a group of generators having the same characteristics behaves exactly the same because the problem is divided into partial problems for each generator. Therefore, at a certain time, the unit group is either in operation of all units or in stop of all units. So, for example,
If the fuel consumption of these units is restricted during the planning period, it is difficult to adjust the fuel consumption.

[0008] In preparing an operation plan using a thermal power generator and a pumped generator, in general, the calculation of the thermal power generator and that of the pumped generator are divided, and after calculating only with the thermal power generator, means for determining the output of the pumped generator is used. Has been taken a lot. However, in this case, it is unclear how much reserve power the pumped generator has (usually a value proportional to the remaining amount of water in the pumped storage reservoir) when calculating the thermal power generator. When a plan is requested, a technician with rich knowledge and experience modifies the initial water level of the reservoir of the PSPP, and repeats trial and error to request the plan. As a result, there is a problem that much labor and time are required.

Accordingly, the present invention has been made in view of the above-mentioned circumstances, and has been developed in consideration of the above situation. It is an object to provide a plan creation device and a recording medium for recording a processing program of the device.

[0010]

According to the first aspect of the present invention, an operation plan for starting / stopping a generator is determined from predicted total power demand information, generator facility data, and operation plan information such as operational constraint information. In the power supply and demand plan creation device to be created, after formulating the operation plan of starting and stopping the generator, which is the supply and demand planning problem, taking into account the constraints that do not extend between generators, the constraint conditions are defined by the Lagrangian relaxation method. A subproblem creating means for relaxing and dividing into subproblems for each generator, and a subproblem for creating a combined subproblem obtained by combining the subproblems divided by the subproblem creating means in consideration of the constraints spanning between the generators A state transition diagram based on dynamic programming based on a synthesizing means and a partial problem and a synthesized partial problem obtained by the partial problem creating means and the partial problem synthesizing means. Operating costs and satisfy the matter is one that has to be provided and the problem solving means to create a management plan to be a minimum. According to this means, with respect to the generator group in which the restriction over the generators is imposed, the partial problem for each generator, which is alleviated by the Lagrangian mitigation method, is reduced to all possible states in each time zone of each generator. It is represented by a state transition diagram to be represented, these transition diagrams are combined, combined, and solved together with a single subproblem. By taking such measures, it is possible to create an optimal operation plan in consideration of the restrictions straddling between generators, which have been difficult to consider in the past.

According to a second aspect of the present invention, in the power supply / demand plan creator according to the first aspect, a state transition of each partial problem obtained by the partial problem creating means and a combined partial problem obtained by the partial problem combining means. In the figure, a limited state transition diagram is created by taking into account the constraint conditions related to starting and stopping, and a solution is obtained. According to this means, in addition to the state transition diagram of each subproblem and the combined subproblem, the constraint condition regarding the start and stop is reflected, and the restriction regarding the start and stop over the generators, for example, the generators in the same power plant Operational restrictions such as the start / stop time difference restriction can be considered.

According to a third aspect of the present invention, in the power supply / demand plan creator according to the first aspect, the state transition of each partial problem obtained by the partial problem creating unit and the combined partial problem obtained by the partial problem combining unit. In the figure, a limited state transition diagram is created by taking into account the range of output that can be obtained from the generator and solved. According to this means, in the state transition diagram of each partial problem and the synthesized partial problem, the range of possible output is limited,
It is possible to take into account operational constraints, such as constraints on generator output that straddles between generators, for example, on upper and lower limits of the total output of a certain power plant.

According to a fourth aspect of the present invention, there is provided the power supply / demand planning system according to the first aspect, wherein a group of generators having the same fuel consumption characteristic as heat consumption characteristics are collectively set as fuel characteristics of one equivalent generator. A means for expressing and calculating the fuel cost characteristic, a means for dividing and solving a partial problem by the Lagrangian relaxation method based on the fuel cost characteristic calculated by this means, and a means equivalent to one generator by this means A means is provided for determining the number of generators operating with the same fuel characteristics by distributing the output to the generators of the generator group again based on the results obtained. According to this means, the generator groups having the same fuel characteristics are represented as one generator equivalent to the generator groups, and then together with the single generator group having the same fuel characteristics, Since the partial problem of is solved, the calculation time can be reduced. Further, for example, if there are six generators having the same characteristics, it is possible to obtain a more detailed plan for starting and stopping than conventional means, such that two of them are operated and the rest are stopped, and operation constraints such as fuel consumption constraints are obtained. Can easily be satisfied.

According to a fifth aspect of the present invention, there is provided the power supply / demand planning system according to the fourth aspect, wherein a group of generators having the same fuel characteristics as heat consumption characteristics are collectively set as fuel characteristics of one equivalent generator. Expressing and calculating a combined fuel cost characteristic, calculating a characteristic boundary of the fuel cost characteristic based on the fuel cost characteristic, and calculating each power generation unit price according to the characteristic boundary calculated by the means. Means and means for presenting data on the number of operating generator groups, synthesized fuel characteristics, characteristic boundaries, and power generation unit prices, and enabling appropriate correction are provided. According to this means, the combined fuel characteristics and the output that is the boundary between the fuel characteristics and the power generation unit price according to the characteristic boundary are calculated, and the calculated value is presented to the operator, and can be corrected if necessary. It is possible to confirm that the fuel characteristics compiled by the user are appropriate, and to correct them if necessary.

According to a sixth aspect of the present invention, there is provided the power supply / demand planning system according to the fourth aspect, wherein a group of generators having the same fuel characteristic as a heat consumption characteristic are collectively equivalent to one equivalent generator fuel characteristic. Means for expressing and calculating the fuel cost characteristic, means for changing the maximum and minimum outputs of the generator group in consideration of the constraint on the total output of the generator groups having the same fuel characteristic, and means for changing this And means for limiting by a state transition diagram so as to satisfy the calculated fuel cost characteristics while satisfying the maximum and minimum outputs changed by the above. According to this means, for a group of generators having the same fuel characteristics, if there is a restriction between generators, the range of output that can be taken by a single integrated generator is limited. Various constraints can be satisfied while using the characteristics.

According to a seventh aspect of the present invention, there is provided a power supply and demand system for generating an operation plan for starting and stopping a generator based on predicted total power demand information, generator equipment data, and operation plan information such as operation constraint information. In the plan making device, a reserve ratio allocation search means for searching for a reserve reserve ratio for each time for the thermal power generator and the pumping generator, a reserve reserve ratio for each time obtained by this means, and an electric power for the total demand power for each time. Based on the demand, a thermal power plant operation plan creating means that creates a thermal power plant operation plan using a mathematical programming method, and the surplus and insufficient power determined by the thermal power plant operation plan created by this means. A pump operation plan creating means for preparing an operation plan of the pumped generator using mathematical programming is provided. According to this means, it is possible to automatically determine the optimal reserve power distribution without requiring human knowledge and experience.

According to an eighth aspect of the present invention, in the power supply / demand planning system according to the seventh aspect, the reserve ratio allocation searching means includes a genetic algorithm, a simulated annealing, or a genetic algorithm and a simulated annealing. The search is performed using both of the above. According to this means, an optimal shared reserve is obtained without requiring knowledge or experience, using a genetic algorithm or simulated annealing as a search method. In general, simulated annealing is known to have a better local search capability than a genetic algorithm.For example, generations are advanced in the processing of a genetic algorithm, and the average fitness of the population is the maximum fitness of the population. Calculate optimal reserve allocation without using human knowledge and experience by using a means that shifts from genetic algorithm to simulated annealing when approaching sufficiently. be able to.

According to a ninth aspect of the present invention, there is provided a power supply and demand system for generating an operation plan for starting and stopping a generator based on predicted total power demand information, generator facility data, and operation plan information such as operational constraint information. On the recording medium that records the processing program of the plan creation device, after formulating the operation plan of starting and stopping the generator, which is a supply and demand planning problem, taking into account the constraints that do not span between generators, A combined partial problem is created by combining the partial problem created by the partial problem creating unit that divides the condition into sub-problems for each generator by relaxing the conditions and the partial problem creating unit in consideration of the constraints spanning between the generators. A partial problem synthesizing unit, and a state by dynamic programming based on the partial problem and the synthesized partial problem obtained by the partial problem creating unit and the partial problem synthesizing unit In shifting diagram in which a recording medium for recording the program in question solving means for creating a management plan that operational costs satisfy the operational constraints is minimized.

[0019]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram of an electric power system supply / demand plan creating apparatus applied to the first to third embodiments of the present invention.

The power supply / demand plan creator for a power system includes a demand forecasting system 1 for predicting the total power demand and other information necessary for creating an operation plan for each time zone, and a total power demand forecast by the demand forecasting system 1. An information transmission device 3 on the system side that transmits information such as demand to the transmission line 2, and an information transmission device on the main body side that receives information such as the total power demand transmitted from the information transmission device 3 via the transmission line 2 Device 4,
A demand and supply plan creation device main body 5 that creates an operation plan with a low operation cost based on the information such as the estimated total power demand received by the information transmission device 4, the generator facility data, and the data on the operational constraints. And a man-machine interface device 6 for inputting necessary instruction information by an operator and displaying a processing result.

Next, the main body 5 of the demand and supply plan creator includes equipment data storage means 11 for storing the maximum output / minimum output of each generator, the minimum stop time of each generator, the minimum operation time, and other necessary data; Based on the supply and demand balance, a partial problem creating means 12 for creating a partial problem for each generator as a partial problem using the Lagrangian mitigation method, and a generator group in which constraints across generators are imposed. And sub-problem synthesizing means 13 for synthesizing the respective sub-problems, and constraint data storage means 14 for storing constraint data such as operational constraints, for example, time difference between start and stop in the same power plant, upper and lower limit values of total output, and the like. And a constraint reflecting means 15 for reflecting these constraint data in the partial problem and the synthesized partial problem.
Solving means 16 for calculating the minimum cost operation plan while satisfying the constraint conditions to be considered and the combined partial problem; total power demand data transmitted from the demand forecasting system 1; And a processing data storage unit 17 for storing result data.

In the demand and supply plan creator 5 constructed as described above, the operation plan is formulated by using a mathematical programming method based on the predicted total power demand, the data relating to the generator, and the operational constraint data. A series of processes to be calculated are automatically executed.

Next, the partial problem creating means 12 which is a main part of the present invention will be described in detail.

First, as shown in FIG. 1, the subproblem creator 12 formulates a generator start / stop plan, which is a supply / demand plan problem, in consideration of constraints that do not extend between generators. The constraints are relaxed by the method and divided into partial problems for each generator.

The Lagrange mitigation method satisfies restrictions such as power supply and demand balance, maximum / minimum output of the generator, minimum operation time of the generator, minimum stop time, start / stop pattern, etc., which do not span between generators. The goal is to find the optimal solution that minimizes the operating cost while doing this.

Specifically, (1) power supply and demand balance,
(2) generator maximum / minimum output, (3) generator minimum stop time, (4) generator minimum operation time, (5) generator start pattern, (6) generator stop pattern, etc. Formulate a generator start / stop plan in consideration of constraints. At this time, considering the objective function from the viewpoint of operation cost, the following (1)
As shown in the equation, the fuel cost and the starting cost are variables. Expression (2) is a constraint.

[0028]

(Equation 1)

Here, in the above equation, T: planning period,
N: total number of generators, D (t): total power demand at time t
Required, P_j(T): generator_jAt time t, P
_jmin:Generator_jMinimum output of P_jmax: Generator j
Maximum output of f_j, S_j:Generator _jFuel cost function, start
Cost function, u_j(T): generator_jLuck at time t
It is in a rolling state (1: operation 0: stop).

Using the objective functions (1) and (2) and introducing a Lagrange multiplier λ (t), a Lagrange function shown in the following equation (3) is created.

[0031]

(Equation 2)

In this equation (3), when the Lagrange multiplier λ (t) is fixed, the problem can be divided into sub-problems for each generator as shown in equation (4).

[0033]

(Equation 3)

Next, the sub-problem solving means 16 shown in FIG.
Is performed in step S16a.

In the processing of the subproblem solving means 16, an operation plan for the subproblem obtained by the subproblem creation means 12 is determined so as to minimize the operation cost when the restrictions imposed by the dynamic programming technique are relaxed. .

Here, the sub-problem was divided into (4)
To solve the equation by a dynamic programming technique, consider a state transition diagram shown in FIG.

In FIG. 3, the circles shown in the figure represent the operating state of the generator, and → represents the transitionable path.

Now, for example, as a route that can transition from the generator in state A (stop time ≧ 3) in the stop pattern J1c at time 1 in the lower part of FIG. 3 to time 2, the start pattern C (output = 50 MW) ) State, and a state B in which the stop is continued as a stop pattern J1c. Subsequently, since the C state is in the middle of the activation pattern, at time 3, it can only transition to D (output = 100 MW) of the activation pattern. next,
At time 4, the transition from the D state can be made to the E state (operating time = 1 hour), and the transition from the E state can be made only to the F state due to the restriction of the minimum operation time of 2 hours.

When transitioning from the F state (operating time ≧ 2) to the stop pattern, either the transition to the H state (output = 60 MW) of the stop pattern or the G state to continue the operation is performed. Can transition to

On the other hand, in the case of the start pattern I state at time 1 (operating time ≧ 2), if an attempt is made to stop, K →
It is necessary to pass a stop pattern such as L → M. The transition from the stop pattern M (stop time = 1) is possible
There are only N states due to the constraint of a minimum stop time of 3 hours.

As described above, in order to calculate the operation cost, the node cost is given to the circle at the time t when the power is generated. In addition, a start cost _Sj is given to the path → which transitions from the stop portion to the start portion. And the minimum operation time
A route from time 1 to time T that minimizes the operation cost is found in consideration of the minimum stop time and the start / stop pattern. Accordingly, when the Lagrange multiplier is fixed, the optimal solution of the partial problem by the following equation (5), that is,
Find an operation plan that minimizes operation costs.

[0042]

(Equation 4)

In general, since the Lagrange multiplier is set to an appropriate temporary value, the constant correction processing S1 shown in FIG.
8 is performed.

That is, the Lagrange multiplier is corrected by the following equation (6) because of the equation constraint (since the power supply and demand balance is not satisfied).

[0045]

(Equation 5)

In the above equation, k is the number of repetitions, a and b
Is a constant.

In the equation (6), the result of the process S16a for solving a partial problem is stored, and when the value of the portion in parentheses [] becomes sufficiently small, the constant correction is ended by the process S18 shown in FIG. .

According to the processing shown in FIG. 2, an operation plan that satisfies the operation constraint and minimizes the operation cost is obtained.

Next, the partial problem synthesizing means 13 will be described.

Restrictions that span generators, for example,
For the generator group to which the time difference at the time of stoppage, the upper and lower limit values of the total output, the minimum number of parallel units, and the like are imposed, the partial problems stored in the facility data storage unit 11 are combined, and the path of the state transition diagram is further added. , A plan is created that satisfies the constraints across generators for these generator groups and minimizes operating costs.

FIG. 4 shows the above processing. First, in step S13a, the intersection of the partial transition state transition diagrams is calculated.
Next, in step S13b, a transition path that transits to each state is created.

As an example, a combined state transition diagram shown in FIG. 7 is created from the state transition diagrams shown in FIG. 5 and FIG.

FIG. 5 shows a transition state of the start / stop pattern of the generator a. In the figure, the horizontal axis shows the time, and the vertical axis shows the start pattern and the stop pattern corresponding to the time. The circle indicates the state and the mark → indicates the possible transition. FIG. 5 substantially corresponds to FIG.

FIG. 6 shows a start / stop pattern transition state of the generator b. In the figure, similarly to FIG. 6, the horizontal axis indicates time, and the vertical axis indicates a start pattern and a stop pattern transition corresponding to the time. A circle indicates a state, and a → indicates a transition possible. FIG. 6 corresponds to FIGS. 3 and 5.

FIG. 7 is a composite of the state transition diagram J2 of FIG. 5 and the state transition diagram J3 of FIG. 6, based on the time representing a plane perpendicular to the paper of FIGS. An example of the display of a cross section, that is, a cross section created at each time is shown.

In FIG. 7, out of the four blocks divided by the dashed lines, J4a indicates the transition of the generator a in FIG. 5 to stop, and J4b indicates the transition of the start of the generator b in FIG. ing. In addition, J4c indicates the stop transition of the generator b in FIG. 6, and J4d indicates the stop transition of the generator a in FIG. J4e indicates the start transition of the generator a in FIG. 5, and J4f indicates the start transition of the generator b in FIG. Further, J4g indicates a stop transition of the generator b in FIG. 6, and J4h indicates a start transition of the generator a in FIG.

As described above, according to the first embodiment, for example, with respect to a unit group in which a restriction is imposed between generators, a partial problem for each generator, which is alleviated by the Lagrangian relaxation method, is described. Combine and combine state transition diagrams that represent all possible states in each time slot,
Solve with other single subproblems. By this means, it is possible to take into account restrictions straddling between generators, which were difficult to consider in the past.

Next, a second embodiment of the present invention will be described with reference to FIG.

In the second embodiment, in addition to the first embodiment, a transition state limitation means 31 for performing a transition state limitation process S31 shown in FIG. 8 is provided.

For example, when a restriction on the simultaneous start / stop time difference is imposed on the generator a and the generator b, the paths of the synthesized state transition diagram are limited, and the restriction is considered.

FIG. 9 is a synthesized state transition diagram J5 in the case of a start / stop time difference 1H between the generator a and the generator b, which is the synthesized state transition diagram J4 described in FIG.
In the state transition diagram J5, when the start / stop time difference between the generator a and the generator b is 1H in the state transition diagram J5, a sufficiently large cost is given in advance as shown by a black circle in the drawing. I have. According to such a state transition diagram, the process S31 shown in FIG.

FIG. 10 is a synthesized state transition diagram J6 in the case of a start / stop time difference 2H between the generator a and the generator b similar to that of FIG. a
In the case where the difference between the start and stop time of the generator b is 2H, a sufficiently large cost is given in advance as shown by a black circle in the figure. According to such a state transition diagram J6,
The process S31 shown in FIG.

As a result, routes can be created in the synthesized state transition diagram by means of the dynamic programming technique so that transition is possible, and the cost of each route is obtained.
Then, a solution that minimizes the cost of each route is obtained,
An operation plan corresponding to the obtained route is required.

As described above, according to the second embodiment, in the state transition diagram of each sub-problem and the combined sub-problem, the restriction on the start / stop over the generators, which is the restriction on the start / stop, for example, the same power generation Operational constraints such as a start / stop time difference constraint on the in-house generator can be added.

Next, a third embodiment of the present invention will be described with reference to FIG.

In the third embodiment, a maximum output / minimum output changing means 41 for performing a maximum output / minimum output changing process S41 is described.
Is provided.

That is, for example, when restrictions on the total output are imposed on the generator a and the generator b, the paths of the synthesized state transition diagram are limited, and the upper and lower limits of the output that can be taken in each state. And a sufficiently large cost is given in advance to a portion that is stopped at the same time in the synthesized state transition diagram. In this way, a solution with the minimum cost cannot pass through these states.

FIG. 12 shows a synthesized state transition diagram J7 in which restrictions on the start / stop patterns of the generators a and b and the total output are given.

In the state transition diagram J7 of FIG. 12, J7a indicates the transition of the generator a to stop, J7b indicates the transition of the start of the generator b, J7c indicates the transition of the stop of the generator b, and J7d indicates the transition of the generator a. Shows the stop transition of. Further, J7e indicates a start transition of the generator a, J7f indicates a start transition of the generator b, J7g indicates a stop transition of the generator b, and J7h indicates a start transition of the generator a. Then, a cost sufficient to satisfy the constraint condition is given to a black circle shown in the state transition diagram J7. In addition, in the state of the free output, in order to satisfy the upper limit value 600 MW of the total output, the processing represented by the equation (7) based on the equation based on the equal λ method is performed.

[0070]

(Equation 6)

By solving the above equation (7), the maximum outputs of a and b are obtained, and the maximum outputs of the new generators a and b can be satisfied to satisfy the upper limit of the constraint on the total output.

As described above, according to the third embodiment, the range of output that can be taken in the state transition diagram of each subproblem and the synthesized subproblem is limited, and the constraints on the generator output over the generators, For example, an operation constraint, which is a constraint on the upper and lower limits of the total output of a certain power plant, can be added.

Next, a fourth embodiment of the present invention will be described with reference to FIG.

As shown in FIG. 13, the demand and supply plan creator includes a characteristic combination calculator 51 for performing a characteristic combination process S51 for combining fuel cost characteristics of a plurality of thermal power generators having the same characteristic into one, and a partial problem creator. , A partial problem solving means 16a for performing a partial problem solving process S16a, and a constant correction process S18 for correcting a fixed multiplier to relax constraints. Means 18 and a number-of-operating-units calculating means 55 which performs a process S52 of calculating the number of operating units which determines the number of generators having the same characteristics by judging the generators having the same characteristics from their outputs.
It consists of

Next, the operation of the characteristic combination calculating means 51 will be described.

For example, suppose that there is a generator whose heat consumption characteristics are represented by the following equation (8).

[0077]

(Equation 7)

The heat consumption characteristic when n generators are operated at the same time is expressed by the following equation (9).

[0079]

(Equation 8)

The characteristic boundary between the heat consumption characteristics of the n-unit operation and the (n + 1) -unit operation is expressed by the following equation (10).

[0081]

(Equation 9)

By solving this, the following equation (11) is obtained.

[0083]

(Equation 10)

Accordingly, the characteristics obtained by combining the units having the same characteristics as described above are as follows when the minimum output and the maximum output per unit are P _min and P _max , respectively.
It is represented by a value approximated by a least square method or the like by a quadratic function of the equation.

[0085]

[Equation 11]

The method of solving a partial problem with respect to a generator whose characteristics are summarized is the same as that of a single generator. After solving a partial problem related to a generator whose characteristics are combined into one unit, means for re-allocating to the original generator is implemented as in the following equation (13). Here, in the following, p is an output obtained as a result of solving a partial problem.

[0087]

(Equation 12)

As described above, according to the fourth embodiment, a group of generators having the same characteristics is expressed as a single generator equivalent to these groups, and together with a group of generators having the same characteristics,
Since it is configured to be solved as a partial problem for each generator, the calculation time can be reduced. Further, for example, if there are six generators having the same characteristics, two of the generators can be operated and the rest stopped, so that a more detailed plan for starting and stopping can be obtained than in the conventional method. Operational constraints such as constraints can be satisfied.

Next, a fifth embodiment of the present invention will be described with reference to FIG.

FIG. 14 is a block diagram of an electric power system supply / demand plan creating apparatus according to a fifth embodiment of the present invention. The same reference numerals as those in FIG. 1 showing the first embodiment denote the same or corresponding parts. Is shown.

The fifth embodiment is different from the fourth embodiment in that the characteristic combination calculating means 51a and the composite characteristic presenting means 5
1b. The characteristic combination calculating unit 51a stores the combined characteristic and the result obtained by calculating the characteristic boundary in the processing data storage unit 17, and notifies the combined characteristic presentation unit 51b. The notified combined characteristic presenting means 51b takes out the characteristic stored in the processing data storage means 17 and the characteristic boundary or the power generation unit price at the characteristic boundary and presents it to the man-machine interface device 6.

FIG. 15 shows power generation unit price tables J8 and J9 calculated from the characteristic boundaries described above.

J81, J82, and J83 of the power generation unit table J8, which are examples of the screen in the figure, correspond to the constants a, b, and c of the equation (8) for obtaining the heat consumption characteristics of the fourth embodiment. Illustrated "5.00E-8""2.0E + 00""3.
3E + 2 ”is a constant a, in the case of one generator A,
These are the numerical values of b and c. Furthermore, the values in italics in the lower column are values of constants a, b, and c when four generators are put together. The horizontal columns for the lower boundary 1 to the lower boundary 5 are (1
The output obtained by the equation (0) is shown, and a column next to the output shows each power generation unit price.

On such a screen, first, constants a, b,
When a numerical value such as c is input, four constants in the italic character portion are calculated, and further, a boundary is obtained, and output values and power generation unit prices corresponding to boundaries 1 to 4 are obtained.
b and c can be changed. Note that the lower J9 has three generators.
It is a screen example in the case of a stand.

As described above, according to the fifth embodiment, the power generation unit price is calculated based on the collected characteristics and the output that is the boundary between the characteristics and the characteristics, and the calculated value is presented to the operator. To make sure that the characteristics summarized by the operator are reasonable,
They can be modified if necessary.

Next, a sixth embodiment of the present invention will be described with reference to FIG.

In the sixth embodiment, the processing shown in FIG. 16 is executed. The characteristic combination calculating means 51 performs processing S51 for synthesizing the characteristics of the generator groups having the same characteristics, and the generator group having the same characteristics. Maximum output / minimum output changing means 41 for performing processing S41 of changing the maximum output / minimum output of the generator with the combined characteristics in consideration of the total constraint of the output with respect to A transition state limiting unit 31 for performing a process S31 for limiting the state of the transition diagram; a partial problem solving unit 16a for performing a processing stage S16a for solving a partial problem;
And a constant correcting means 18 for performing a constant correcting process S18.

First, in the process of FIG. 16, the maximum output of the generator is set as the upper limit of the total output, the minimum output is set as the lower limit of the total output, and the transition state is limited by the process S41 of the maximum output / minimum output changing means 41. By the processing S31 of the means 31,
A large cost is given in advance to the stop portion of the state transition diagram of the generator. Next, in the process S16a for solving a partial problem, a solution is obtained in which the stopped state cannot be obtained and the upper and lower limits of the total output cannot be obtained in the free operation. Therefore, the constraint on the total output can be satisfied.

As described above, according to the sixth embodiment, when a generator group having the same characteristics has a limitation that extends between generators, the range of the output that can be taken by a single generator equivalent to the restriction is set. Since the limitation is made, various restrictions can be satisfied in a state where the collected characteristics are used.

Next, a seventh embodiment and an eighth embodiment of the present invention will be described with reference to FIG.

The illustrated demand / supply plan creation device main body 5 includes equipment data storage means 11 for storing the maximum output / minimum output of each generator, the minimum stop time of the thermal power generator, the minimum operation time and other necessary data, Means for searching for the optimal allocation ratio of the reserve reserve to the thermal power generator and the pumping generator in the country using a genetic algorithm (GA), and mathematically based on the power supply / demand balance based on the total power demand. A thermal power plant operation plan creating means 23 for creating an operation plan of the thermal power generator using a planned method, and a pumping generator using a mathematical planning method for surplus and insufficient power determined from the operation plan of the thermal power generator. Pump operation plan creating means 24 for creating the operation plan of the power plant, and a processing data storage unit for storing total power demand data and processing result data transmitted from the demand forecasting system 1 side. It is provided and 17.

Here, the reserve ratio allocation search means 22 is provided with
The process shown in FIG. 8 is performed, and the gene creating means 71 performs a process S71 of expressing the ratio of the reserve reserve of each of the thermal power unit and the pumping unit at each time by, for example, a gene using a binary number. Next, a process S72 of calculating the fitness of each gene created based on the process S71 by the gene creating means 71 is performed by the fitness calculating means 72. Further, after the calculation of the degree of matching, a process S73 of rearranging the genes and manipulating the genes of the next generation is performed by the process S73 of the gene manipulating means 73. Further, the convergence determining means 74 performs a process S74 of determining whether the manipulated gene has advanced to a predetermined maximum generation. In this determination, when the process has progressed to the maximum generation, it is determined that convergence has occurred. Subsequently, the processing S75 of selecting the operation plan of the thermal power plant and the water pump corresponding to the gene having the highest degree of matching from the genes remaining in the largest generation when the convergence is determined by the processing S74 of the convergence determining means 74 is optimal. This is performed by the plan selection means 75.

As an embodiment of the thermal power plant operation plan creating means 23, there is a means based on a mathematical programming method such as the Lagrangian relaxation method described above. For example, in the Lagrange mitigation method, it is difficult to consider constraints expressed by inequalities such as reserve constraints, so when creating an operation plan, among generators that are not in parallel at the time when reserve reserve is insufficient. And means for arranging the generators with the lowest power generation unit price in parallel.

The pump operation plan creating means 24 performs the processing shown in FIG. 19, and performs the processing shown in FIG. 19 on the surplus electric power, for example, by the pumping allocating means 81 which performs processing S81 of allocating pump water from an efficient generator. . Next, a process S82 of calculating the amount of water pumped by the pumping operation is performed by the used water amount calculating means 82. Then, for example, the processing S83 of allocating the pumped water to the generator by the equal lambda method using the used water amount characteristic is performed by the pumping allocating means 83. afterwards,
This is performed by the reserve power calculating means 84 which performs the processing S84 of calculating the reserve power based on the amount of the water storage device or the like. In this case, when the target reserve capacity has not been reached, the reserve capacity increase processing means 85 performs a process S85 of pumping water for the purpose of increasing the remaining water level.

Next, the spare ratio distribution search means 2 shown in FIG.
The operation of No. 2 will be specifically described.

First, the gene creating means 71 creates a gene for designating the share ratio of each of the thermal power unit and the water pump unit at each time when the pre-price rate is set. For example, when the gene at a certain time is “0011001111”, the sum of the number 1 is set as the ratio of the heating power. That is, the thermal power of the reserve is set to 6/10, and the pumped water is set to the remaining several 4/10. Such a gene is created for each time.

In this case, in order to reduce the number of genes deviated from the optimal plan, an initial population of genes is first prepared based on the calculated reserve ratio, and a reference individual is created. On the other hand, using random numbers, sequentially create bit inversions of several places by the preset number,
The creation of the initial population is completed when the created number of individuals reaches a preset number of individuals.

FIG. 20 is a flowchart of the process for calculating the reserve for gene creation in step S71 shown in FIG.

First, processing S91 is performed on the total demand power obtained from the demand forecasting system or the like by the thermal power equipment operation plan creating means 91, and an operation plan for only the thermal power equipment is created. As a result of the preparation of the thermal power plant operation plan, surplus / deficiency calculating means 92
The surplus supply power and the shortage of supply power are calculated by the process S92 according to. A pump operation plan is prepared for the surplus supply power and the shortage of the supply power by using the process S93 by the pump operation plan creating means 93.

Next, after the load distribution of the pump is determined, the thermal power sharing load calculating means 9 is obtained by subtracting the pumped water from the total demand.
4, and is calculated in step S94. On the other hand, the operation plan of only the thermal power unit is calculated again by the process S91a by the thermal power unit operation plan creation unit 91. As a result, processing 8
The reserve power of each of the thermal power unit and the water pump is calculated by 5a. Based on this, a reference gene is created based on the definition of the gene described above.

Next, the adaptability calculating means 72 shown in FIG. 18 prepares each gene based on the result of creating each operation plan using the reserve capacity of each of the thermal power unit and the water pump unit represented by each gene as a target value. Is calculated. This degree of conformity is calculated using, for example, equation (14). In the genetic algorithm (GA), the degree of conformity is expressed by a reciprocal in order to perform maximization.

[0112]

(Equation 13)

Next, the gene manipulation means 73 for performing the process S73 shown in FIG. 18 rearranges the genes to create the next generation of genes. Genetic manipulation includes growth, crossover, and mutation. As in the recombination example J9 shown in FIG. 21, the propagation produces the same child as the parent (upper row in the figure). At that time, the number of individuals having a higher degree of fitness should be increased. Intersection replaces part of the two parents to create two children (middle shown).
Mutation refers to changing a part of the parent to create a child (lower row in the figure).

Next, the convergence determining means 74 for performing the processing S74 shown in FIG. 18 determines whether or not the generation has advanced to a preset maximum generation. If the generation has not reached the maximum generation, the convergence degree shown in FIG. The process returns to the processing S72 of the calculating means 72. If the process has proceeded to the maximum generation, a process S79 is performed by the optimal plan selecting means 79 shown in FIG. The optimal plan selecting means 79 sets the operation plan corresponding to the gene having the highest degree of matching among the genes remaining in the largest generation as the optimal operation plan.

As described above, according to the eighth embodiment, by applying the genetic algorithm, it is possible to search for the optimal allocation of the reserve, and as a result, all the constraints to be considered in operation are satisfied and the operation cost is reduced. A low-cost operation plan can be created automatically without manual trial and error.

Next, an electric power system supply / demand plan creating apparatus according to an eighth embodiment of the present invention will be described with reference to FIG. 17 and FIG.

The power system supply and demand plan creation apparatus according to the eighth embodiment is simulated as a search means in the search for the optimal reserve of thermal power and pumping water at each time by the body of the supply and demand plan creation apparatus in FIG. The method of annealing is applied.

Here, Simulated Annealing (hereinafter referred to as “S”)
A), which is one of the optimization methods applicable to the combinatorial optimization problem, and is a method of simulating “annealing” on a computer.

In general, ordered crystals of a substance are obtained by melting them at an elevated temperature and then cooling it down slowly enough. In this phenomenon, the state with the minimum physical energy corresponds to the regular crystalline state, and when the energy is minimal, the crystalline state does not become regular. Annealing is a physical means for obtaining a regular crystalline state as a global minimum without being trapped by a minimum value of energy by giving a fluctuation of temperature to a system. SA has been proposed as a solution to an optimization problem such as a combination optimization problem by analogy with this physical “annealing”. That is,
SA is a stochastic optimization method for finding a global minimum of a nonlinear function having many local minima.

The reserve ratio allocation search means 22 specifically performs the processing shown in FIG. 22, and calculates the ratio of the reserve power shared by the thermal power generator and the pumped generator at each time.
For example, distribution plan creating means 1 represented by a character string using a binary number
01, the process S101 is performed. The processing S102 of calculating the evaluation value of the operation plan created based on the distribution plan created by the distribution plan creating unit 101 is performed by the evaluation value calculating unit 10
2 Next, processing S102 of changing the distribution plan to a new distribution plan is performed by the distribution plan changing means 103.
Next, in step S104, the evaluation value after the change is calculated.

Further, the distribution plan judging means 104 for judging whether or not to adopt the created new distribution plan is subjected to processing S
104 is performed. Further, the processing S105 is performed by the convergence determining means 105 which controls the change of the distribution plan and repeats the processing up to a predetermined condition. This convergence determination means 105
When it is determined that the distribution plan has converged, the optimal plan selection unit 107 performs a process S107 of selecting an operation plan corresponding to the distribution plan with the best evaluation value from the distribution plan history up to that time.

First, the distribution plan creation means 101 of the reserve ratio distribution search means 22 creates a distribution plan of the ratio of the reserve reserve of the thermal power plant and the water pump at each time. Now, regarding the creation of the distribution plan, for example, the number of 1s in a character string consisting of 10 digits of 0 and 1 indicates the ratio of the reserve power shared by the thermal power unit. Specifically, for example, the necessary reserve power at a certain time is 1000 MW, and the character string is “001110101”.
If "1", 6 out of 10 characters are 1, so
The thermal power plant will bear 60% of the necessary reserve power. That is, the total reserve power of the thermal power unit is 600 MW. This is stored in the processing data storage means 17.

Next, the evaluation value calculation means 102 gives an evaluation value to the operation plan created based on the reserve capacity distribution plan. This evaluation value is calculated using, for example, equation (15).

Evaluation value f = operation cost * 10E-6 + penalty (15) where operation cost = fuel cost + startup cost penalty = number of constraints violating the operation constraints

After calculating the evaluation value by the evaluation value calculating means 102, the distribution plan changing means 103 creates a new distribution plan by, for example, randomly selecting one portion of the character string and inverting 0 and 1. I do.

The distribution plan judging means 104 calculates the difference Δ between the evaluation values before and after being changed by the distribution plan changing means 103, for example.
f is calculated and, for example, the following (1) is calculated as the difference Δf between the evaluation values.
In the case of using equation (6), Δf <0, that is, a new plan is adopted for improvement. In the case of Δf> 0, that is, in the case of deterioration, a new correction plan represented by the probability r (17) is adopted. That is, the plan before the change is maintained with the probability (1-r).

[0127]

[Equation 14]

T is a value called a temperature parameter, and this parameter is initialized to, for example, a predetermined initial temperature TS.

The convergence determining means 105 determines, for example, whether or not the temperature parameter has reached a predetermined minimum temperature TE. When it is determined that the temperature parameter has reached the minimum temperature TE, the procedure shifts to the optimal plan selecting means 107. Convergence determination means 1
05 indicates that, for example, a change in the revision plan at a constant temperature is 1
If it is determined in advance to be 0 times, it is determined whether or not the correction plan has been changed up to a maximum of 10 times. If the correction plan has been changed up to the maximum number of times, the temperature is lowered according to, for example, the following equation (18).

T (N + 1) = 0.9 * T (N) (18) where TE ≦ T (N) ≦ TS

The optimum plan selection means 107 is a distribution plan having the best evaluation value in the change history of the distribution plan until the specified number of times is changed at the lowest temperature (the smallest one when the evaluation value is the expression (15)). Is selected as the optimal operation plan.

The simulated annealing is as follows.
It is known that local search ability is superior to genetic algorithm. For example, in the processing of a genetic algorithm, when the generation progresses and the average fitness of the population approaches the maximum fitness of the population sufficiently, the genetic algorithm can be switched from the genetic algorithm to simulated annealing. A more optimal reserve allocation than using alone can be determined without the need for human knowledge or experience.

Next, a ninth embodiment of the present invention will be described with reference to FIG.

In the embodiment shown in FIG.
U132, a main memory 133, and a transmission bus 134, a computer 131, and a general-purpose SCSI cable 1
It comprises a recording medium writing device 135 and a recording medium 136 connected by 37 or the like.

In such a configuration, the main memory 133 in the computer 131 has the processing program according to the invention described in the first to eighth embodiments. Such a program is stored in the recording medium writing device 13.
5 is arbitrarily recorded on the recording medium 136. Another computer is the recording medium 136 or the recording medium 136
The processing program can be executed by using another recording medium written from the computer.

As described above, according to the embodiment of the present invention, by solving the partial problem alleviated by the Lagrangian relaxation method and the problem obtained by arbitrarily combining them, the supply-demand balance and the generator are not crossed. Not only restrictions,
Constraints that span generators can be considered.

Also, by combining the characteristics of the generator groups having the same characteristics, a detailed operation plan of the generator groups having the same characteristics can be created.

[0138] Further, since the optimal reserve power to be shared by each of the thermal power unit and the water pump is searched by a genetic algorithm or SA, all the constraints to be considered in operation are satisfied and the operation cost is low. A practical operation plan can be created automatically and quickly.

[0139]

As described above, according to the first aspect of the present invention, with respect to a group of generators that are restricted over the generators, the state transition of the partial problem for each generator, alleviated by the Lagrangian relaxation method, is performed. This is represented by a diagram, and these state transition diagrams are combined and solved together with a single subproblem, so that it is possible to create an optimal operation plan taking into account constraints that straddle generators, which were difficult to consider in the past. it can.

According to the second aspect of the present invention, in addition to the state transition diagram of each partial problem and the synthesized partial problem, the restriction condition regarding the start and stop is reflected, and the restriction regarding the start and stop over the generators, For example, it is possible to consider the operation restriction of the start / stop time difference restriction regarding the generators in the same power plant.

According to the third aspect of the present invention, in the state transition diagram of each partial problem and the combined partial problem, the range of possible outputs is limited, and the constraint on the generator output extending between generators is defined. For example, operational restrictions such as restrictions on the upper and lower limits of the total output of a certain power plant can be considered.

According to the fourth aspect of the present invention, a generator group having the same fuel characteristics is expressed as a single generator equivalent to those generator groups, and then a generator group having the same fuel characteristics is obtained. Simultaneously, the partial problem for each generator is solved, so that the calculation time can be shortened.Moreover, since the number of generators operating with the same fuel characteristics is set, it is possible to obtain a more detailed plan for starting and stopping than conventional means. Therefore, it is easy to satisfy operational constraints such as fuel consumption constraints.

Further, according to the invention of claim 5, the combined fuel characteristics, the output at the boundary between the fuel characteristics, and the power generation unit price based on the characteristic boundary are calculated, the calculated value is presented to the operator, and corrected if necessary. Since it is possible, it is possible to confirm that the fuel characteristics compiled by the operator are appropriate, and to correct them if necessary.

According to the sixth aspect of the present invention, when a generator group having the same fuel characteristics has a limitation that extends between generators, the range of output that can be taken by a single generator equivalent to the restriction is set. Because of the limitation, various restrictions can be satisfied while using the collected fuel characteristics.

According to the seventh aspect of the present invention, the optimal reserve power distribution can be automatically obtained without requiring human knowledge and experience.

According to the invention of claim 8, an optimal sharing reserve is obtained without requiring knowledge and experience by using a genetic algorithm or a simulated annealing as a search method. Furthermore, a more optimal reserve allocation using the genetic algorithm alone can be obtained without requiring human knowledge and experience.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a power system supply and demand plan creation device according to a first embodiment of the present invention.

FIG. 2 is a flowchart showing processing of a partial problem and a combined partial problem solving means provided in the power supply / demand plan creation device of FIG. 1;

FIG. 3 is a state transition diagram showing the operation of the partial problem solving means of the first embodiment.

4 is a flowchart showing a process of a partial problem synthesizing means provided in the power supply and demand plan creating device of FIG.

FIG. 5 is a diagram showing a first example of a state transition diagram of a generator before synthesis.

FIG. 6 is a diagram showing a second example of a state transition diagram of the generator before synthesis.

FIG. 7 is a diagram showing a synthesized state transition diagram.

FIG. 8 is a flowchart showing processing of a transition state limiting unit according to the second embodiment of the present invention.

FIG. 9 is a diagram showing a limiting means in the case of a one-hour stop time difference on the synthesized state transition diagram.

FIG. 10 is a diagram showing a limiting means in the case of a two-hour stop time difference on the synthesized state transition diagram.

FIG. 11 is a flowchart showing processing of maximum output / minimum output changing means according to the third embodiment of the present invention.

FIG. 12 is a diagram illustrating a limiting unit on a state transition diagram synthesized when there is a restriction on total output.

FIG. 13 is a flowchart showing a process performed by the power supply and demand plan creation device according to the fourth embodiment of the present invention.

FIG. 14 is a configuration diagram of a power supply and demand plan creation device according to a fifth embodiment of the present invention.

FIG. 15 is a diagram showing a power generation unit price table calculated from characteristic boundaries.

FIG. 16 is a flowchart showing the processing of the power supply and demand plan creation device according to the sixth embodiment of the present invention.

FIG. 17 is a configuration diagram of an electric power system supply / demand plan creation device according to seventh and eighth embodiments of the present invention.

18 is a flowchart showing a process of a reserve power distribution search means provided in the power supply / demand plan creation device of FIG.

19 is a flowchart showing a process of a pump operation plan creating means provided in the power supply and demand plan creating device of FIG. 17;

FIG. 20 is a flowchart showing a reserve force calculation process.

FIG. 21 is an explanatory diagram showing an example of rearrangement by genetic manipulation by a genetic algorithm.

FIG. 22 is a flowchart showing the processing of reserve power distribution search means according to the eighth embodiment of the present invention.

FIG. 23 is a diagram showing a ninth embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Demand forecasting system 2 Transmission line 3, 4 Information transmission device 5 Demand-supply plan creation device main body 6 Man-machine interface device 7, 24, 93 Pumping machine operation plan creation means 11 Equipment data storage means 12, 52 Partial problem creation means 13, 53 Partial problem synthesizing means 13b Transition bus creating means 14 Constraint data storing means 15 Constraint condition reflecting means 16, 64 Partial problem solving means 17 Processing data storing means 18 Constant correction means 20 Operation plan creating means 21 Equipment data storing means 22 Reserve ratio allocation Searching means 23, 91 Thermal power plant operation plan creating means 31 Transition state limiting means 41 Maximum output / minimum output changing means 51a Characteristic synthesis calculating means 51b Synthetic characteristic presenting means 54, 65 Constant correction means 55 Operating number calculation means 62 Maximum output / minimum Output modifying means 63 State limiting means 71 Gene Creation means 72 Fitness calculation means 73 Gene manipulation means 74, 105 Convergence determination means 75, 79 Optimal plan selection means 81 Pumping allocating means 82 Water usage calculating means 83 Pumping allocating means 84 Reserve power calculating means 85 Reserve power handling processing means 92 Surplus / deficiency calculation means 94 Thermal power sharing load calculation means 101 Distribution plan creation means 102 Evaluation value calculation means 103 Distribution plan change means 104 Distribution plan determination means 107 Optimal plan selection means 131 Computer 132 CPU 133 Main memory 134 Transmission bus 135 Recording medium writing Device 136 Recording medium 137 Cable

Continuing from the front page (72) Inventor Masahiko Isuzu 1 Toshiba-cho, Fuchu-shi, Tokyo Inside the Toshiba Fuchu Plant, Inc. Person Seiichi Kato 1 Toshiba-cho, Fuchu-shi, Tokyo (72) Inside the Toshiba Fuchu Plant Co., Ltd. No. 1, Toshiba-cho, Toshiba-shi, Fuchu Plant, Toshiba Corporation (72) Inventor, Masaru Hiromasa No. 1, Toshiba-cho, Fuchu-shi, Tokyo GB06 HA14 KA03 KA16 KC06 KC10 KC12 KC23 KD67 LA16

Claims (9)

[Claims] 1. A power supply and demand plan creation device for creating an operation plan for starting and stopping a generator from predicted total power demand information, generator equipment data, and operation plan information such as operation constraint information, After formulating the operation plan of starting and stopping the generator, which is a supply and demand planning problem in consideration of the constraints that do not span between generators, the constraints are relaxed by the Lagrangian relaxation method to reduce the partial problem for each generator. A partial problem creating means for dividing, a partial problem synthesizing means for creating a combined partial problem by combining the partial problems divided by the partial problem creating means in consideration of the constraints spanning between the generators, and the partial problem creating means And a state transition diagram obtained by dynamic programming based on the partial problem obtained by the partial problem synthesizing means and the synthesized partial problem. And a problem solving means for creating an operation plan with a minimum cost. 2. In the state transition diagram of each partial problem obtained by the partial problem creation means and the partial problem synthesized by the partial problem synthesis means, a state transition diagram limited by taking into account a constraint condition regarding starting and stopping. The power supply and demand plan creation device according to claim 1, wherein the solution is created and solved. 3. In the state transition diagram of each partial problem obtained by the partial problem creating means and the partial problem synthesized by the partial problem synthesizing means, the range in which the output obtained from the generator can be taken is limited. 2. The power supply and demand plan creation device according to claim 1, wherein the created state transition diagram is created and solved. 4. A means for calculating a fuel cost characteristic by collectively expressing a group of generators having the same fuel characteristic as heat consumption characteristic as an equivalent fuel characteristic of one generator, and calculating a fuel cost characteristic by the means. Means for dividing and solving into partial problems by the Lagrangian relaxation method based on the cost characteristics, and based on the result solved as one equivalent generator by this means, the output is again output to each generator of the generator group. 2. A power supply and demand plan creating device according to claim 1, further comprising means for determining the number of generators operating with the same fuel characteristic by distributing the same. 5. A group of generators having the same fuel characteristic as heat consumption characteristics are collectively expressed as fuel characteristics of one equivalent generator, and a combined fuel cost characteristic is calculated. Means for calculating a characteristic boundary of the fuel cost characteristic based on the characteristic boundary; means for calculating each power generation unit price according to the characteristic boundary calculated by this means; the number of operating generator groups, the combined fuel characteristic, 5. The power supply and demand plan creation device according to claim 4, further comprising means for presenting the data of the characteristic boundary and the unit price of power generation and enabling the data to be appropriately modified. 6. A means for calculating a fuel cost characteristic by expressing a group of generators having the same fuel characteristic as heat consumption characteristics as a fuel characteristic of an equivalent single generator, and a generator having the same fuel characteristic. Means for changing the maximum and minimum outputs of the generator group in consideration of the total output constraints in the group; and the calculation while satisfying the maximum and minimum outputs changed by the change means. 5. The power supply and demand plan creation device according to claim 4, further comprising means for limiting the power consumption characteristic by a state transition diagram so as to take into account the calculated fuel cost characteristic. 7. A power system supply and demand plan creating apparatus for creating an operation plan for starting and stopping a generator based on predicted total power demand information, generator equipment data, and operation plan information such as operational constraint information, A reserve ratio allocation search means for searching for a reserve ratio for each time for the thermal power generator and the pumping generator, based on the share reserve ratio for each time obtained by this means, and the power demand for the total demand power for each time, A thermal power plant operation plan creating means for preparing an operation plan of a thermal power generator using a mathematical programming method, and a mathematical programming method is used to compensate for surplus and insufficient power determined by the operation plan of the thermal power generator created by this means. And a pump operation plan creating means for creating an operation plan of the pumped generator. 8. The reserve ratio allocation search means includes a genetic algorithm, a simulated annealing,
The power supply and demand plan creation device according to claim 7, wherein the search is performed using both a genetic algorithm and simulated annealing. 9. A process of a power system supply / demand plan creation device for creating an operation plan for starting and stopping a generator based on predicted total power demand information, generator facility data, and operation plan information such as operational constraint information. In a recording medium for recording a program, after formulating an operation plan for starting and stopping a generator, which is a supply and demand planning problem, in consideration of constraints that do not span generators, the constraints are relaxed by a Lagrangian relaxation method. Means for dividing into sub-problems for each of the generators, and a sub-problem synthesizing means for generating a synthesized sub-problem by synthesizing the sub-problems divided by the sub-problem creating means in consideration of the constraints spanning between generators Means, and a state by dynamic programming based on the partial problem and the composite partial problem obtained by the partial problem creating means and the partial problem synthesizing means. A recording medium for recording a processing program with a problem solving means for creating an operation plan that satisfies operation constraint conditions and minimizes operation cost in a transition diagram.