Hydraulic Engineering Software IV

Sewerage Systems Spielvogel, J. Edenhofer

h of

(tk), ... ,a ( ... R}

-·,-v --"·"'

Allocation Fairbairn Reservoir, Queensland, Australia K.S. Tickle (*), I.C. Goulter (**)

Control System Y. Wang, S. Wang, X. Su

Systems Design G.C. Santana (*), S. Soares (**)

dimensions and imply operational diameters. The optimal design of water distribution system installation operational costs subjected limits in operational conditions. Yates et al. established an analysis about computational complexity in minimizing water supply does mean that there instances, depending on physical and initial datas used, which difficult to solve using any specific algorithm. solutions presented in the literature are divided into classes. The first that reduces is not easy to different algorithms system because the problem has multiple local solutions. this present an algorithm to the water distribution system design problem based nonlinear programming problem. algorithm uses

operational variables f(d,h) subjected head node vector (a head piezometric level obtained by the elevation pressure); is the pipe velocity vector, and is the set of diameters. cost function includes

tendency to violate head flow velocity bounds cost solution, penalty method used to enforce the limits. In order to avoid ill-conditioning of simple penalty functions (Bertsekas [2]), an Lagrangian penalty function yielding

should be noted that the alqorithm determines the reduced qradient analytically, resultinq in less computational effort than other that use numerical approaches (6]). Tests have shown that the accuracy is the discrete diameter variables

leads higher flow velocities. Table 1. Solutions for two flow velocity bounds. Pipe Relaxed Discrete 1.04 1.10 1.04 1.00 0.54 0.60 0.49 0.50

in the modified layout does not occur terminal point, as expected. This to the influence of the elevation that can yield a hiqh enerqy level in the least pressure (the critical node). In fact, althouqh is the critical node, the water flow direction from node to 4. includes a •critical concept" that reduces the computational effort avoidinq the explicit determination of the supply heads. technique allows determine the exact optimal solution for the

is declared basic. if the in first optiaization the large reservoirs, to set of priori tlzed

easily be stored iteration to the just redefining the pointer. This flexibility in

historical records. priorities are

N.-1. Win, G.L. Vandewiele

S.H. AbuBakar (*), P.J.B. Morrell (**), A. Potts (**)

Transmission C. Popovska, B. Todorovski

, where is the average if coefficient is defined

Improvement by Implementing Autonomous Acquisition Network in Roy (*), R. Gauthier (*), M. Hetu (**)

Relationship Stations- Calculation Method in HYDROS and STOQ H.J. Henriksen

Reduced Model J. Polo (*), L. Castillo(**), J. Armengou (*), J. Dolz (*)

Control in a Hydrometric Network M. Baldin (*), M. Gonella (**) performs the useful tests (correlation analysis, double analysis, etc.), deterministic part, that is the object of this paper, based on results of numerical simulations carried out alonq the river.

in the Branch there is the confluence the Sangone River and in the Branch the confluence of the Panaro River. The changements of discharges due to the significant variations of hydrometric levels the values are acceptable because of the singularities effects or there are failures. calculation of the theoretical differences between water levels, due to

gA~ + A=cross surface level above the datum; distribution factor; q•discharge of the inflow; x=streamwise distance and t•time. one-dimensional implicit, numerical scheme

present Branches. At the the procedure analyses the possible to apply to control contemporary data real time. procedure analyses yearly time series: for each day of the year the hydrometric level is recorded. time step in data recording

Confidence Using and SQD ( procedure calculates the confidence limits [4]: ,k) ± n SQD(i,j ,k) n=l,3 , (8) distribution of the values DIF[i,j,d] in the confidence intervals. This calculation each reference stage [k] and gives an image of the

calculate average parameters of the distributions: the results of the controls are general indicators of significance. deterministic control procedure presented this paper permits analysis of single data and help to find out any existing can be interfaced with an automatic system data acquisition. In this is possible to carry out real time control of measured water levels. 1. Danish Hydraulic Institute,

System for the Purpose of Real-Time Flood Forecasting A. Cabal, M. Erlich

J.C. Lauria

hi msel ook ng at. huge mass of numbers lhout. begin from. Therefore, lhe challenge develop syst.emat.ic procedures for conceptual analysis so as lhe pattern of numerical results. stepwise proposal for describing t.he fundamen- tals of valve application here.Afler of works characterizing lhe subject., diment.ionless intended lo describe lhe operating behaviour of valves installed elements introduced. It. has long been recognized t.he importance of valve flow characlerislic in predicting steady stale

can be divided valve loss all singular es valve premises basic case. ng available head Reynolds- are a•neric as whole. Speci exper alone, are combined

specif'ic as angle, or area. def'ined as rat.io generic one, so as have f'ixed range 1. Generalized diagrams Kv-x-ar curve f'or valve

probable valve Several appear original scope generalized diagrams. earring on idea cases valve smaller ed pumping- 4. 5] Reynolds number i uence generalized diagrams the basic case. these topics are treated here. the be pleased to send request.

range. usual lo determine lhe operating pcdnls by means of guidelines or in nslances, in case-by-case basis. The former, al besl, is proce- dure surrounded by doubt because of lhe uncertainly lo the significant resemblance of the the previous one thal generated lhe rule. latter possible there are conditions for controlling lhe In addition, the valve-closure schedule can be specified according lo the pallern of lhese curves: lhe lower lhe variation, lhe faster lhe closure- element movement may be, and vice versa.

lations. success computers in conlribuling beller design easier and operation slart.s lh an applying basic principles lhal transcend individual methodologies.

L.D. Fagherazzi, J.-C. Rassam, L. Carballada

seasonal using models whose parameters typically vary from season to season. In the second approach, IOdels incorporating disaggregation procedures sites of interest, and then disaggregate them into seasonal flows. of the characteristics of t .ast appropriate to the ob of the study. a x 52) matrix of weekly flows of a river with m years of record. basic idea of the proposed .athod is that the principal of the 52 variables represented by the of are uncorrelated and therefore they can be generated independently. Using the singular decomposition

Matrix of spaulated

the number of seasons. The completion module can used to estimate missing data and/or extend the record necessary. A generalized maintenance of variance extension technique [2,9] is used. the data using the set of display functions the historical record and/or the synthetically generated sequences. the at-site seasonal means,

with those obtained from the historical record. The statistics for comparison include seasonal statistics such as standard deviations, skewness coefficients, month to correlations and lag-zero cross-correlations. validation module computes an additional run Figure 6 illustrates the correlation between successive weeks using the 1950-1990 historical series and the generated series. Figure 7 the historical lag-0 cross-correlations the generated ones.

properties. The modular structure of the code and user friendly menus will widen the use of stochastic hydrology thus improve water resources operation at

2. J.C., J.R. and H.B., (1989), 'A Generalized Maintenance of Variance Extension Procedure for Correlated Series', Water Resour. Res., 25(3), 345- 3. J.C., and J.R., (1990), A

Chen, R.A. Falconer

A. Gabos

efficient rainfall-runoff models that take into

stat. sections

to assess the feasibility of the fixed pattern of demand. last option the user to the data results either in a tabular or graphic fora. Graphics plotter.

Generation Section Groisman

conformal orthogonal transformations are The two main approaches are: a) methods based on the solution algebraic complete survey of grid generation techniques presented in Refs. [1] to [5]. one, although ensures univalent mapping and grid naturally smooth, requires computational effort similar to that for the solution of equations. The algebraic approach, on the other very attractive because the time chosen because simplicity for or parabolic problems, the latter being the motivation for the investigation.

1\ (t) Equation ( 3) we observe that the interpolant functions must be continuously differentiable order less than the level of desired for the coordinates. If in Equation are chosen

plane of symmetry. typical cross section in Fig.(2) as well as the plane of symmetry of the duct. The cartesian system of reference the shape of the cross section of the duct, for the flow will calculated, looks like suitable to use ellipsoidal coordinates as curvilinear coordinates. of an ellipsoidal coordinate system used for ellipse in Fig. (3). Here the transformation from plane to plane x-y or vice- versa can be calculated by known analytical functions.

Fiq. 4 CUrvilinear coordinate system Points A, F and K must be included as data for each cross section. The number of points for each cross section can be different. It is convenient that each cross section specified by the qiven points is as plain

have with and of given points in the envelope curve. that the focal distance point Fig.(6) Fittina of the envelope the intermediate control 1-m[

successful irriqated aqri culture of any developinq country like India. Irriqation systems, comprisinq of a number of hydrological variables that are stochastic in nature, are differinq place place and time to time, thus influencinq the quality successful when all the operations are par with the existinq environment. in the irriqation environment occur rapidly to chanqes in natural-eco systems activities. chanqes create environment under which an irriqation system difficult to acclimatize the farminq practices under existinq conditions. Therefore expert systems qreat potential for providinq decision-sup- port to irriqation systems farmers.

uncertain information for decision- are ac- quired through experience. Expert systems essential- ly consist of the following Acquisition possible for the developer capture the knowledge and preserve expert system code. base which contains the specific consisting of simple facts, rules that describe the relations, characteris-

expert in . cotton crop management on the basis of atmospheric information. this paper, expert system that has been de- veloped for planning and managing irrigated cropping system expert system consists of two phases: in the phase, an expert system for crop selection has been developed and in the second phase, crop management decisions are included in the expert system for each crop that has selected in the phase. the management actions followed during goes on varying continuously over time and space. To substantiate the effect of climate, crop raised in the previous year without any insect-pest attack, not be successfully grown in the current year as the climate may be conducive for the growth of insects and pests. Similarly the

proper manurinq, proper time and amount, are the most important decisions the manaqement repectively. While irriqation schedules are influ- enced mainly by weather variables such as rainfall, temperature, evaporation etc, alonq with availabili- ty of water plant response, be based on the type of crop and explained in the following sections. formulation and conceptualization In the problem formulation st users and domain expert and ob identi- fied. For the problem-Crop selection, the role of the variables involved in selecting crop

pesticides, storaqe and market- qoverninq variables were identified to select crop. For scheduling irrigation water, the variables identified Rainfall Evapotranspiration furnish the values of variables to be used in the the experts. The information reqardinq crop characteristics adoptability different environment have been collected expert. The aqronomist was consulted qated crop manaqement and the knowledqe was extract- ed through interviews with him by the use of ques- tionnaires. After several sessions with the expert, the values of the variables to be used in the study have been obtained. This form of raw information was

furnished information was converted of production rules expert system developed by Five Research was used for representinq the knowledqe on with simple IF-THEN type of also has ability to perform mathematical operations, provides interface with written other lanquaqes. This feature was useful to update the forecasted values of rainfall with the actual rainfall to schedule irriqation water accordinq to

regard been prepared for each category of crops. 222 rules developed phase and in the second phase, 484 rules were developed for recommendinq water and fertilizer schedules. created rule-base was implemented on a personal testing validating the rules. field data input into the system and the deci- provides recommendations on technical and agronomi- in the second phase for each crop that

Activate ETP 20cm Activate RAIN of crop : Release 4.2 em at a frequency of 6 days

Alonso (*), J. Cuena (*), B. Reig (**)

Software for Hydroelectric Systems J .M. Bourgueil, B. Blervaque

fori=l

Approach for Fluid Transients Analysis Ouazax (*), A.R.D. Thorley (**), B. Akdi (*), Yz zogh (*)

et are approximated by: ,i•1,2,

entities that

illustrates the overall amsnt-··· Root------l

ita vector r.

d' Ingtnieurs, Intl

Knowledge Bases for Real Time Flood Management J. Cuena (*), M. Molina (*), L. Garrote (**)*

Modelling W.G. Wright (*), A.M. Schulte (**)