JPS60122405A - Control system of power generating water flow of connecting water system - Google Patents

Abstract

PURPOSE:To simplify the control by the intervention of the operator by computing the optimum control calculation of the entire water system including all dams and power generating stations of the water system in high speed altogether so as to attain the optimum control in following to the dynamic change in the water system. CONSTITUTION:Plural power generating dams 1-3 are arranged in series from the upper stream to a river and reservoirs 4-6 are provided corresponding to the dams 1-3. Moreover, power generating stations 7-9 are installed to the dams 1-3 and power is generated by water stored in the reservoirs 4-6. Furthermore, water ways 13-15 are connected to the dams 1-3 so as to give ineffective effuluent to the reservoirs 4-6. Moreover, a power generating gate 16 and a direct effuluent gate 17 are provided to the dams 1-3 respectively and setting value controllers 18, 19 are installed to them. The controllers 18, 19 controls the gates 16, 17 so as to discharge set water flows W1, W2. The controllers 18, 19 are provided with a gate opening converter 24 converting the set water flow 20 into a gate opening 21 and a gate driver 25 outputting a motor drive signal 22 by using the opening 21 and gate opening information 23.

Description

[Detailed Description of the Invention] [Field of Application of the Invention] The present invention uses predicted values of the water storage level of each dam in a connected water system and the amount of inflow water from each dam catchment area (hereinafter referred to as a mountain stream) up to a certain time. Based on this, in order to minimize the amount of ineffective discharge of the entire water system, this paper relates to a system for controlling the amount of water generated in a connected water system, which allows the amount of water generated at each power plant to be set at high speed and controls the power generation gate to respond to dynamic changes in the water system.

[Background of the invention]

We previously proposed the following method to control the amount of water generated in a connected water system. In other words, in this method, on the day before the control is performed, each dam is calculated based on the predicted water storage volume at 0:00 of each dam and the predicted value of the mountain stream for 24 hours on that day, and after satisfying the specified water storage volume at 24:00 on that day. Dam control was carried out according to the amount of generated water planned for each dam to minimize the amount of ineffective discharge.

Errors in mountain stream prediction values were mainly absorbed by the buffer capacity of the reservoir, or adjusted by experienced operators by manually changing the amount of generated water and gate discharge. In addition, if the error becomes too large, the planned value is corrected by correction calculation on the day, and thereafter control is performed according to the corrected value. This control method requires (1) a long calculation time, and online control of changes in the water system is not easy; For this reason, the planned value often becomes meaningless. (2) Since the amount of water generated and the amount of gate discharge are determined sequentially from upstream for each dam, it is not possible to minimize the ineffective discharge of the water system as a whole.

(3) Since prediction errors are manually corrected, there are many problems such as the inability to optimize the entire water system, and a new control method is desired from the perspective of effective use of water as an energy resource. It is rare.

Another option is to use linear programming to solve (2) invalid discharge from the perspective of the entire water system, but the computer capacity required to solve it is large and the calculation takes time, so online control is I don't like it. Under the present circumstances, where rainfall is difficult to predict and we cannot expect any sudden improvement in the accuracy of water flow prediction into water systems, the above water system calculation methods have the serious drawback of not being able to reflect dynamic changes in water systems. I have no choice but to think about it.

[Objective of the Invention] The present invention solves the above-mentioned problems, integrates all dams and power plants in a water system, and uses a means to perform optimization control calculations for the entire water system at extremely high speed. The purpose is to provide an optimal control method that follows dynamic changes. Furthermore, another objective is to create a model that is easy for an operator to understand, to facilitate operator intervention, and to provide a system that facilitates reflection of human know-how in calculation results.

[Summary of the invention]

A feature of the present invention is that in a connected water system, the natural phenomenon of water flowing from upstream to downstream, sometimes with a time delay, and in other cases almost without a time delay, We focused on the fact that power plants in a water system can be represented as nodes and can be expressed very naturally in a spatio-temporal network that connects the nodes with waterways, and by using a high-speed solution method called network planning, we can calculate the amount of operation for control. ,
The aim is to make sequential decisions and optimize water use in response to changes in the water system from time to time. Furthermore, by handling a type of nonlinearity that cannot be handled by linear programming, it has the advantage that more optimal control of the amount of generated water can be performed in consideration of the efficiency of the generator. this is,
Determining the amount of water to be generated 3- involves non-linear considerations, such as the fact that the maximum efficiency of the generator is achieved when the amount of water is slightly below the maximum amount of water, and that the penalty for ineffective discharge increases as the amount of water increases. This facilitates the handling of this type of problem, which often requires handling of gender. The present invention can be summarized as follows.

(1) The entire water system is expressed as a spatiotemporal network in which nodes called dams are connected by branches called waterways.

(2) Provide the current water storage volume of each dam and the predicted mountain stream value for a certain period of time to the source node of the network.

At the same time, a cost function that indicates the characteristics of that branch is set for each branch of the network.

(3) Determine the minimum cost flow of the network.

(4) Set the network flow rate corresponding to the current moment in each gate control device as the generated water amount and gate discharge amount.

(5) Each gate control device performs the process 4- cormorant.

[Embodiments of the invention]

Hereinafter, the present invention will be explained in detail with reference to Examples. As an example, consider controlling the amount of power generated in the connected water system shown in FIG. In the same figure, there are dams 1 and 1 on the same river. Dam 2. Dam 3 is installed in series from upstream, with reservoirs 4 and 5 respectively.
．． 6. Each dam has seven power plants.
, 8.9 are installed, and the water stored in the reservoirs 4, 5.6 is used to rotate the generator to generate the required power. It is assumed that water from rainfall in each dam catchment area flows into each dam through routes 10, 11, and 12. In addition, 13°, 14.1.5 respectively indicate the coastal road, and the invalid discharge water from the dam gate is
It is assumed that the water will flow along this seaway and reach the downstream dam. It is assumed that the water that has passed through power stations 7, 8, and 9 reaches the reservoir of the lower dam in a time that can be ignored in calculations, and that the water that flows through seaways 13, 14, and 15 reaches the reservoir of the lower dam after Δ time. It is assumed that there is an inflow.

Each dam has a power generation water gate 16 as shown in Figure 2.
and a dam direct discharge gate 17, and a set value control device], 8.19 is installed for each. This set value control device 18, 19 controls the amount of water W1. The purpose of this is to control each gate to release W2.
The configuration is as shown in the figure. In FIG. 3, the water amount/gate opening converter 24 converts the set water amount 20 into a gate opening 21, and the gate driving device 25 compares the gate opening 21 with gate opening information 23. and opening degree 2
1 is achieved, the motor drive signal 22 is output. Now, in the above-mentioned connected water system, the main part of the present invention will be described, which determines the set amount of water to be supplied to the power generation water gate 16 and the dam direct discharge gate 17 in response to the constantly changing state of the water system. As mentioned above, the present invention is characterized by expressing a connected water system as a network. An example of a network representation of the connected water system shown in FIG. 1 is shown in FIG. This figure is a planning network for 2Δ up to the time ahead. This network will be explained in detail. Number nodes in the network correspond to dams. Alphabet nodes are additional nodes for calculations. Examples of nodes arranged horizontally constitute a spatial network at the same time, and nodes arranged vertically constitute a temporal network at the same spatial position.

For example, nodes 101, 102, and 103 are at time 0°Δt
, 2Δt, represents the state of dam 1, and node 101
, 201, 301 indicate the states of dams 1, 2.3 at time 0. In this calculation model, the control unit time is set to Δt, which is made to coincide with the coastal flow time. The branches connecting each node represent waterways. For example, branch 401 is at time 0
(control time), this is the branch through which water is discharged from the power generation gate at the power plant 1 and used for power generation. Branches 401-4
09 are branches through which water for power generation flows, all similar to the above. From the above assumption, it is assumed that water for power generation reaches the downstream dam without any time delay, so branches 401 to 409 all connect nodes 7- at the same time. Branches 501 to 506 are branches expressing the amount of water stored in the dam. For example, branch 501 indicates the amount of water stored in dam 1 at time Δt. Branches 801 to 803 also indicate the amount of water stored in the dam, and these are branches that flow the amount of water stored in each dam at time 0, that is, the amount of water stored in the dam calculated from the current observed water level. Branches 601 to 606 are branches through which the amount of water discharged from the dam direct discharge gate flows, and since the seaway has a time delay of Δ, they are diagonal branches such as nodes 101→202. This corresponds to the fact that water discharged from dam 1 to seaway I3 at time 0 flows into dam 2 at time Δt. 701 to 703 are the inflow amounts 10 and 11 shown in FIG.
．． This is a branch that represents the number 12. 901 to 908 are additional branches for calculation, and the network is a multi-input multi-output network,
This is for converting one person's power into one output. The additional nodes are for convenience of calculation; A is a node for supplying or drawing the water storage amount of each dam at time 0, H is a node for supplying or drawing the water storage amount of each dam at the calculation end time to the network, and B is a node for supplying or drawing in the water storage amount of each dam at time 0.

C, D are routes 10, 11, 12 to 8- at each time. A node, E, supplies an inflow of , to the network.

F and G are nodes that absorb the amount of outflow to the seaway 15. Further, S and T are called all-time sources and all-time sinks, and are additional nodes for converting a multiple-input multiple-output network into a single-power single-output network.

The above network is created using the following steps.

(1) Create nodes equal to the number of dams in the water system.

(2) Set branches 401, 402, and 403 indicating water for power generation.

(3) Create a time source 89 time sink E, set 701, 702, 703 from B to each node, and set 4
Connect 03 to E.

(4) Repeat the same procedure for the planning period. That is, 1
01-303. B-G nodes and 401-409.70
Branches 1 to 709 are set.

(5) Setting branches 501 to 509 and 601 to 609.

(6) Create nodes A and H, set branches 801 to 803 from A to each dumb node, and at the same time connect branches 507 to 509 to H.

(7) Create all-time source S and all-time sink T, and set necessary branches 901 to 908.

As can be seen from the above method, in this method, branches are set only between nodes through which water flows, so when using a computer, the memory capacity can be greatly reduced. (compared to linear programming).

For each school in the network shown in FIG. 2, the maximum flow rate flowing through that branch and the cost incurred by water flowing through that branch are added. This cost was set as a function of flow rate, and the basic idea is as follows.

(1) If flowing water to that branch produces a profit, reduce the cost of that branch. Branch 401 showing water for power generation
409 corresponds to that.

(2) Increase the cost of that branch if allowing water to flow through that branch will result in a loss. Gate discharge 601~
606 corresponds to that.

(3) Describe the cost setting method in detail for the branch where the designated flow rate is to flow from 0 to less. First, power generation water 401~
The cost function of 409 is shown in FIG. This is a non-linear function such that the cost of power generation at the highest efficiency point is the lowest, but this method has the advantage of being able to handle such non-linearity.
There is one feature.

The cost function for the stored water 501-506 is as shown in FIG. 6, and the cost function for the gate discharged water 601-606 is as shown in FIG. Also, 701-709, 801-803, 901-9
08 is the specified flow rate, that is, in the case of 701 to 709, the predicted value of the stream flow (the predicted value is determined by the operator), and in the case of 801 to 803, the current water storage volume of each dam must be discharged. , and the cost in this case is the 8th
It will look like the figure. The specific value of each cost may be determined and set by the operator as appropriate. For example, 402 than 401
If you want to flow more water (you want to make the power generation amount of the power plant 1 at time 0 greater than that at time Δ1o), the cost of 402 may be made smaller than the 11-cost of 401.

The connected water system shown in Figure 1 is now expressed as a network,
We were able to set a cost function on that branch. Although this example is simple, it can be seen that it can be easily extended to more complex cases. Furthermore, since the specific setting of costs is left to the operator, there is an advantage that the plan intended by the operator can be easily obtained. This, and the fact that calculations are fast, makes the operator invisible (the optimum value has been hit, but the operator does not know how it was reached, and the resulting (impossible to repair) can be removed.

The network representation method of a connected water system, which is one of the key points of the present invention, has been shown above. Next, an example of a solution method for this network will be shown. A cost is attached to the above network, and the planned value at each time is obtained by calculating the minimum cost flow of this network. Various methods have been proposed for solving the network minimum cost flow, and any of them may be used. Here, we consider the current quantity as a fast solution1.
2- Explain the primal dual method. The feature of this algorithm is that the calculation starts from a state where all branch flow rates are 0, and the operation of finding the minimum cost route and the maximum flow on that route is repeated until all the specified flow rates given are satisfied. That is what it is. That is, perform the following procedure.

(1) Set the flow rate of all branches to 0.

(2) Find the least cost path from all-time source S to all-time sink T.

(3) Allocate the maximum flow rate that can flow through the above route to each school on the route.

(4) Continue the procedures in (2) and (3) above until all branches with the specified flow rate flow at the specified flow rate.

For details on the above primal dual method, please refer to books (System Optimization Theory, Kiyotaka Shimizu, Coronasha, etc.).

With the above procedure, when the flow rate of each school in the network is determined, the branch at time 0, which is the current time, is 401°4.02,4
03 and 601, 602, 603 are dam 1.
Dam 2. This is the power generation gate water amount from dam 3 and the dam direct discharge amount. Therefore, it is sufficient to send this to the set value control devices 18 and 19 above. An example of a system for controlling the amount of water generated in a connected water system using this network solution method is shown in Fig. 9. In FIG. 9, 26 is a mountain stream prediction device that calculates a mountain stream predicted value 32 that is input to the network calculation device 28. These 26 functions may be replaced by an operator. 27 is a water level/water storage amount conversion device that calculates the current storage amount 31 of each dam reservoir based on the observed water level from each dam reservoir 4, 5.6. Reference numeral 28 denotes a network calculation device, which is a computer capable of performing the network calculation described above. This first sets the data 31.32 from 26 and 27 as the designated flow rates of branches 801-803 and 701-709, respectively, and performs the above-mentioned least cost method calculation.

The execution results 33 are sent to the set value control device of each dam, respectively.
It is sent in the next response and is used for control.

401 → 1.18402 → 218 403 → 318 601 → 1.19601 → 219 603 → 319 However, 11.8, 218, 318 are power generation gate water volume set value controllers 119, 219, 319 are direct discharge gate water volume set value controllers It is. Moreover, since this calculation is sufficiently fast, it is also possible to display the calculation result to the operator via the display device 34, obtain his approval, and send the output 33. If the operator needs to change the results, he can use the keyboard 35 to adjust the costs for each school in the network.

〔Effect of the invention〕

As explained above, according to the present invention, by only providing the predicted value of the mountain stream and the current water level, the amount of generated water and dam gate discharge amount that minimizes the amount of invalid discharge from each dam can be instantaneously calculated and controlled. be able to.

This results in the following effects.

(1) Control can be easily corrected for errors in predicted values of mountain streams, and control measures that match the dynamics of water systems are provided.

(2) It is possible to minimize the ineffective discharge of the water system as a whole15- (3) It is easy for the operator to intervene in the calculation results,
It removes the invisibility of control. In addition, network representations are easy to understand because they fully correspond to concrete objects such as dams and waterways.

(4) Unlike current calculation methods, long-term predicted values are not required, and the optimum solution at that point in time can be determined only by relatively short-term prediction, so control performance can be improved.

[Brief Description of the Drawings] Fig. 1 is a diagram showing the configuration of a connected water system as an example of the application of the present invention, Fig. 2 is an explanatory diagram of a dam gate operation method, and Fig. 3 is a diagram showing dam gate control. Simple configuration diagram of the device, No. 4
The figure is a network representation of the connected water system in Figure 1, and Figures 5 to 8 are branches 401 to 409 that flow water for power generation in Figure 4, and branches 501 to 5o6 that indicate dam storage water, respectively.
．． Branches 601 to 606 showing gate discharge. Figure 9 shows an outline of the cost functions added to branches 701 to 709, 801 to 803, and 901 to 908 with designated flow rates, respectively. It is a figure showing an example of composition. OI M Spear 2 Figure 3 Figure Oki 5 Zuon 6 Figure Penetration Arisui! E 7 Figure θ Buni i support amount + 8 Figure

Claims (1)

[Claims] 1. In a connected water system consisting of multiple power generation dams installed on a river and one or more power plants installed at each dam, the current water level observation value of each dam and the collection of data for each dam. Based on the predicted amount of inflow water from the water area, the current amount of power generation water for each dam and the amount of water discharged from the dam gate required to extract the maximum amount of power generation from the entire water system are determined at high speed, and the above-mentioned amount for each dam is determined. A connected water system power generation water amount control method characterized by being equipped with a water amount control device that controls the amount of water used for power generation and the amount of water discharged.