Docstoc

OPTIMIZATION OF RESERVOIR OPERATION USING NEURO-FUZZY TECHNIQUES

Document Sample
OPTIMIZATION OF RESERVOIR OPERATION USING NEURO-FUZZY TECHNIQUES Powered By Docstoc
					  International Journal of Civil Engineering OF CIVIL ENGINEERING AND
  INTERNATIONAL JOURNAL and Technology (IJCIET), ISSN 0976 – 6308
  (Print), ISSN 0976 – 6316(Online) Volume 4, Issue 2, March - April (2013), © IAEME
                            TECHNOLOGY (IJCIET)
ISSN 0976 – 6308 (Print)
ISSN 0976 – 6316(Online)
Volume 4, Issue 2, March - April (2013), pp. 149-155
                                                                           IJCIET
© IAEME: www.iaeme.com/ijciet.asp
Journal Impact Factor (2013): 5.3277 (Calculated by GISI)               © IAEME
www.jifactor.com




      OPTIMIZATION OF RESERVOIR OPERATION USING NEURO-
                      FUZZY TECHNIQUES

                                          S. K. Hajare
                           Principal, Someshwar Polytechnic College,
                           Someshwarnagar, Tal. Baramati, Dist. Pune


   ABSTRACT

           Water resource engineering is mainly concerned with planning, designing and
   operation of water resources system. This paper explores the use of soft computing tools for
   water availability and operation studies, deciding suitable advanced optimization techniques
   for understanding a real life study related to water resources engineering and extension of
   these techniques to reservoir operation studies

   Keywords: Water resource engineering; soft computing; optimization; ANN.

   I. INTRODUCTION

          Water resource Engineering is mainly concerned with planning, designing and
   operation of water resources system. Uncertainty in availability of water in space and time
   poses challenges for efficient planning and design of water resources systems. The basic
   techniques used in water resources systems analysis are optimization and simulation, where
   optimization techniques are meant to give global optimum solutions and simulation is a trial
   and error approach leading to the identification of the best possible solution. Optimization
   models are characterized by a mathematical statement of the objective function and a formal
   search procedure to determine the values of decision variables for optimizing the objective
   function (Gupta and Gupta Amit, 1994). The principal optimization techniques are,
   A) Linear programming
   B) Nonlinear programming
   C) Dynamic programming.


                                               149
International Journal of Civil Engineering and Technology (IJCIET), ISSN 0976 – 6308
(Print), ISSN 0976 – 6316(Online) Volume 4, Issue 2, March - April (2013), © IAEME

        Water resources development is a production process and the purpose of
production is to convert a set of inputs to set of outputs, e.g. output for water resources
are irrigation hydropower generation and flood damage alleviation and the examples of
inflow are natural stream flow. A reservoir operating policy is a sequence of decisions in
operational periods (such as months) specified as a functions of the state of the
system.The state of the system in a period is generally defined by the reservoir storage at
the beginning of the period and the inflow to the reservoir during the period (Vedula and
Mujumdar, 2005)Once the operating policy is known the reservoir operation can be
stimulated in time with a given inflow sequence. Traditionally reservoir operation is
based on heuristic procedures, embracing rule curves and to certain extent subjective
judgment by the operator. As such it does not take into account the randomness or
stochasticity of reservoir inflow and rainfall in the irrigated area, inter -seasonal
competition for water among multiple croops.The conventional methods are based on
mass curve analysis and time series analysis and have limitations such as.

   • Paucity of data and quality of data can be major impediment for reliability of
     results.
   • Inability to handle or manage a complex tasks under significant uncertainty.
   • Assumes a definite sequence of events as in rainfall data but it is subjected to
     considerable time variations
   • In many approaches the structures and parameters of the model usually do not
     have any physical significance.
   • Dose not considers the inception by the recognition that the human brain taking
     into account such as nonlinear modeling, classification association and control.

    In order to find remedies for such limitations it needs to explore use of soft
computing as well as advanced optimization tools and techrliques such as artificial neural
networks ,fuzzy logic, Neuro-fuzzy etc.

II. POLICES FOR RESERVOIR OPERATION

2.1 Standard Operating Policy

        The standard operating policy (SOP) aims to best meet the demands in each period
based on the water availability in that period. It thus used no foresight on what is likely to
be the scenario during future period in a year. Let D and R represent respectively the
demand and release in a period. Let the capacity of the reservoir be K.Then the standard
operating policy for the periods is represented in fig.1. The available water in any period
is the sum of storages S at the beginning of the period and inflow Q during the period.
The release is made as per the line OABC on the Fig.1




                                             150
International Journal of Civil Engineering and Technology (IJCIET), ISSN 0976 – 6308
(Print), ISSN 0976 – 6316(Online) Volume 4, Issue 2, March - April (2013), © IAEME




                               Fig.1: Standard operating policy

Along OA : Release = water available; Reservoir will be empty after release.
Along AB : Release = demand; Excess water is stored in the reservoir (fill up phase)
At.A      : Reservoir is empty after release.
At.B     : Reservoir is full after release.
Along BC : Release = demand + excess of availability over the capacity (spill)

       The release in any time period is equal to the availability ,S + Q or demand, D
whichever is less as along as the availability, does not exceed the sum of the demand and the
capacity. The standard operating policy no optimization criterion is used in the release
decisions, for highly stressed systems standard operating performs poorly.

III. OPTIMAL OPERATING POLICY

        One of the classical problems in water resources systems modeling is the derivation of
an optimal operating policy for reservoir to meet a long term objective. Modeling techniques
to be used depend on whether the reservoir inflows are treated deterministic or stochastic. As
shown in Fig.2 given single simplified reservoir system of known capacity K and sequence of
inflows.




                               Fig2: Single Reservoir Operation

       The reservoir operation problem involves determining the sequence of release Rt that
optimizes an objective function. In general the objective function may be a function of the
storage volume and /or the release.
                                              151
International Journal of Civil Engineering and Technology (IJCIET), ISSN 0976 – 6308
(Print), ISSN 0976 – 6316(Online) Volume 4, Issue 2, March - April (2013), © IAEME

3.1 Rule curves for reservoir operation

        A rule cure indicates the desired releases or storage volume at given period of the year
in a steady state condition, some rules identify storage volume targets that the operation is to
maintain, as far as possible and others identifies storage zones ,each associated with a
particular release policy. The rule curves are best derived through simulation for a specified
objective, although for simple cases it may possible to derive them using optimization.

IV. LITERATURE REVIEW

4.1 Linear programming

       In reservoir operation linear programming (LP) is well known most favored
optimization techniques as it is easy to understand and does not require any initial solution.
For optimal reservoir operation, objective function is to maximize net benefits of the crop
subjected to various constraints such as storage capacity, canal capacity, Area under crop,
Evaporation losses Seepage losses, Inflow and its dependability. Demand of water for various
uses such as industrial, water supply, irrigation etc. An intraseasonal allocation model can be
used to maximize sum of relative yields of all crops, for given state of system by using LP
number of models developed on LP such as integrated model for optimal reservoir operation
developed deterministic LP. Model for short term annual operation (Long Le Ngo,
2006).Limitation of LP is that it cannot consider stochastic as well as random nature of
inflow and demand.(Vedula and Nageshkumar,1996).

4.2 Dynamic Programming /Real Time Reservoir Operation

        Dynamic Programming (DP) is a sequential or multistage decision making process
works on a divide and conquer manner. it gives the steady state operational policy of single
reservoir by using backward recursive equations. Steady state models are useful in deriving
polices for maximizing long-term benefits from irrigation system. The steady state operation
model developed by (Vedula and Muzumdar, 1992 ) focuses on main bases for real time
operation model and it is better in case of critical low flow years. It is useful in finding
steady state policy selecting stage as a time period in a particular year (after which it is
assumed that reservoir is no longer useful).The real time operation is formulated for solution
ones at the beginning of each inter seasonal period. It uses forecasted inflow for the current
period and all subsequent period in the year. (Muzumdar and Ramesh, 1997)

4.3 Limitations

 i) Assumption that the decision made at one stage is dependent only on the state variables
     and is independent of the decision taken in other stages.
 ii) The dynamic programming will not be an appropriate technique where decision is

4.4 Evolutionary Algorithm (EA)

       Real world reservoir problem mostly involves complexities like discrete, continuous
or mixed variables, multiple conflicting objectives, non-linearity, discontinuity etc. for such a
                                              152
International Journal of Civil Engineering and Technology (IJCIET), ISSN 0976 – 6308
(Print), ISSN 0976 – 6316(Online) Volume 4, Issue 2, March - April (2013), © IAEME

situation stochastic search as EA is used. EA provides not only a single best solution but 2nd
best 3rd best and so on as required ,also gives quick approximate solution and may be
incorporated with other local search alogorithms.To handle limitations of EA and the
convergence It is required to carry out problem oriented sensitivity analysis to find out the
range in which model is effective. (Reddy Janga &Nagesh kumar,2006).

V. ARTIFICIAL NEUTRAL NETWORKS

      Artificial Neutral Network (ANN) is structured to resemble the biological neutral
network

Two aspects:

i) Knowledge acquisition through a learning process and
ii) Storage of knowledge through connections known as synapade weights.
        The artificial neutron has two modes of operation ,the training mode and using mode
in The training mode, the neutron can be trained to fire (or not) for particular input patterns.
When a taught input pattern is deleted at the input, its associated output to become the
Current output (Zurada,2006). Artificial neural networks are also used successfully for Single
reservoir as well as multireservoir operation(Raman and chandramouli1996). ANNs are
particularly useful as pattern recognition tools for generalization of input Output
relationship. In the water recourses engineering most common application of ANNs
Includes thus for rainfall runoff relationships, stream flow forecasting and reservoir
Operation (Vedula and Mujumdar,2005). For reservoir operation the ANN is trained by the
available in flow with related to out flow At various demand of water. Then it can be
applicable for the stochastic nature of inflow as Well as demand.

VI. FUZZY LOGIC CONTROL

          Fuzzy logic is extension of classical set theory and element is the member of several
sets. At the different degree. Fuzzy sets are defined by labels and membership functions. The
Fuzzy rues will be relies on human experts to express knowledge of appropriate control
Strategies (Ross, 1995) .Many successful applications of fuzzy systems were reported in the
literature especially in control and modeling (Fontana etai, 1997). The main advantage of the
fuzzy control method is to control the processes that are too Complex to be mathematically
modeled. Reservoir operation mode fuzzy logic haveFollowing distinct steps (Vedula and
mujumdar, 2005).
   i) Fuzzification of inputs where the crisp in puts such as the inflow, reservoir Storage and
      release are transferrin to fuzzy variables.
   ii) Formulation of fuzzy set, based on a expert knowledge base.
   iii) Application of a fuzzy operation to obtain one number representing the premises of each
       rule.
   iv) Shaping of the consequence of the rule by implication and defuzzification.

       The fuzzy neural network (FNN) model has been developed to study the behavior of
optimal release operating policies formulated through DP (Deka and Chandramouli,2009)

                                              153
International Journal of Civil Engineering and Technology (IJCIET), ISSN 0976 – 6308
(Print), ISSN 0976 – 6316(Online) Volume 4, Issue 2, March - April (2013), © IAEME

6.1 Integration of Models

         The challenge of stochastic DP locates that it is usually associated with difficulties in
data availability and solution efforts, especially when there are more thAn two state variables
(Ravikumar and Venugopal,1998)The Fuzzy provides a convinient approach for optimizing
complex reservoir system with low quality information where the uncertainty cannot be
represented adequately by probability theory (Liu,2011). so that a methodology by
integration chance constrained programming and factorial design is used to account for
uncertainties in reservoir operation and management. The impact of system reliability and
initial reservoir storage as well as their interaction are examined through a set of factorial
design experiments and complex interrelationships among water release, initial reservoir
storage and reliabilitylevel are examined through implementing the developed chance
constrained model based on variety of system condition (Li,2003)

6.2 Suggested Methodology

       From the literature review some of typical problem associated with operation for
various methods are as indicated below.

GAPS
     • LP cannot consider stochastic as well as random nature of inflow and demand.
     • DP/ Real time reservoir operation assumes that the decision made at one stage is
         dependent only on the state variables and is independent of the decision taken in other
         stages but in reservoir operation decision taken at one period effects on the other.
     • For EA techniques problem oriented sensitivity analysis should be carried out to find
         out the range in which model is effective.
     • SDP is associated with difficulties in data availability and solution efforts when there
         are more than two state variables.
  Hence it is suggested to explore the following options:-
Consider water users in different seasons under different reliability levels and initial storage
conditions may emphasis on design of cropping pattern according to seasonal as well as
initial availability of water. Explore use of soft computing techniques for reservoir operation
and management under uncertainties by effectively relating the information quality of system
variable to the reservoir operation decisions. The physical system may be single reservoir or
multi-reservoir with stochastic nature of inflow to meet demand.

VII. EXPECTED OUTCOMES

        Explore the use of soft computing tools for water availability and operation studies.
Deciding suitable advanced optimization techniques for understanding a real life study
related to water resources engg. and extension of these techniques to reservoir operation
studies. Use of human brain intelligent to solve large complex problem in water resources
engineering by use of rational decision in a environment of uncertainty. Attempt for
application to real life problem involving a water resources system in the state of Maharashtra
state.


                                               154
International Journal of Civil Engineering and Technology (IJCIET), ISSN 0976 – 6308
(Print), ISSN 0976 – 6316(Online) Volume 4, Issue 2, March - April (2013), © IAEME

REFERENCE

  [1] Deha Paresh Chandrs and Chandramouli V.(2009) Fuzzy Neural Network Modeling of
      oir Operation ASCEJ.Water Resource .Plng.Mgt,135(1),5-12.
  [2] DmtanceD.G.,Gates T.K. and Moncada,E.(1997) Planning reservoir operation with
      imprecise objectives.J.Water Resour Plng.Mgt,123(3)432-510.
  [3] Gupta B.L. and Gupta Amit (1994) water resource system management standard pub.
      Distributor;new Delhi.
  [4] L.J.B.(2003)Integration of stochastic programming and factorial design for
      optionalreservoir, Operation,ISEIS-J. Environmentalinformation,1(2),12-17.
  [5] Ltu,B(2001)Fuzzy random chanced-constrained programming,IEEE Trans fuzzy
      system 9 (5),713-720.
  [6] Long Le Neg(2006).Optimizing reservoir operation (Ph.D.Thesis),A case study of
      Binnreservoir,Vietnam (Institute of Environment and Resources) Technical University
      Denmark.
  [7] Mujurmdar P.P.and Ramesh T.S.V.(1997) Real Time reservoir operation for irrigation
      ASCE, J. Water Resour Plann.Manag122(5),342-347.
  [8] Ravikumar V.and Venigopal k.(1998)Optimal reservoir operation using mulit
      objective evolutionary algorithm Springer J,Water Resour,Manag 20(1)86-8.8.
  [9] Ross.T.J.(1995)Fuzzy logic with engineering application Tata McGraw Hit New
      Delhi.
  [10] Vedula S and Mujumadar P.P.(1992).Optional reservoir operation for irrigation of
      multiple crop,J.Water REsour.Res.28(1)1-9.
  [11] Vedula S and Nageshkumar D.(1996)An integrated model for optimal reservoir
      operation for irrigation of multiple crops,water Resource.Res,32(4),1101-1108.
  [12] VeddiaS.and Mujumadar P.P.(2005)Water resources system, modeling technique and
      analysis, Tata McGraw-Hill, New Delhi.




                                          155

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:1
posted:4/18/2013
language:
pages:7