Çok makinalı güç sistemlerinde parametre adaptif kontrol yönteminin incelenmesi
Investigation of parameter adaptive control method for MMPS
- Tez No: 39181
- Danışmanlar: PROF.DR. M. EMİN TACER
- Tez Türü: Doktora
- Konular: Elektrik ve Elektronik Mühendisliği, Electrical and Electronics Engineering
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 1993
- Dil: Türkçe
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 121
Özet
ÖZET Bu tezde, çok makinalı güç sistemi içindeki turbo- generatörlerin uyarmaları ve mekanik girişlerinin birlik te kontrolü adaptif kontrol tekniği ile simüle edilmiş tir. Son yıllarda adaptif kontrol tekniklerinin uyarma kontroluna uygulanması önem kazanmıştır. Bu çalışmada, uyarma ile mekanik regülatör girişlerinin adaptif kontro lü, çok giriş-çok çıkışlı biçimde, güç sistemindeki her makina üzerinde uygulanmıştır. Kullanılan Parametre Adaptif Kontrolü modelinde, seçilen kestirim algoritması işlem sayısı daha az olan Projeksiyon II kestirimi olup, hesaplama süresinde kazanç sağlamıştır. Tezdeki çalışmalar beş bölüm altında toplanmışlardır. Bu bölümler kısa başlıklar halinde aşağıdaki gibidir. Birinci bölümde, adaptif kontrolün güç sistemleri ne uygulanması konusunda literatür araştırmaları ve konuya kısa bir giriş yapılmıştır. İkinci bölümde, adaptif kontrolün en önemli kısımlarından birini oluşturan kestirim algoritmaları tanıtılmış, Projeksiyon II testirimini kullanan ve daha önce tek makinalı sisteme uygulanmış olan adaptif yöntem tek giriş- tek çıkışlı sistemler için tanıtılmıştır. Üçüncü bölümde, sistemin modeli oluşturulmuştur. Bunun için, önce sistemi oluşturan senkron generator, uyarma, mekanik regülatör, transformatör şebekeye ait modeller kurulmuş ve bunlar için varsa sınırlamalar belirtilmiştir. Aynı bölümün daha sonraki ayrıtlarında, kısa devre durumu ve sonrasındaki şebeke değişimlerinin nasıl modellendiği anlatılmıştır. Dördüncü bölümde, güç sistemlerinin kontroluna duyulan ihtiyaç ve kullanılan adaptif algoritmanın tezde uygulanan şekilde çok giriş-çok. çıkışlı durumu tanıtılmış, simülasyon incelemesinin alt programlara dayanarak gerçekleştirilmesi sunulmuştur. Tezde turbogenerator birimleri ve şebeke için verilen ilk değerler ile hesaplanan sürekli çalışma büyüklükleri ve simülasyon sonuçlarını gösteren eğriler yine bu bölümde verilmiştir. Beşinci bölümde, tezde elde edilen sonuçlar verilmiş, gelecekte yapılması önerilen çalışmalardan bahsedilmiştir. -xıı ?
Özet (Çeviri)
SUMMARY INVESTIGATION OF PARAMETER ADAPTIVE CONTROL METHOD FOR MMPS In the thesis, the application of adaptive control ler to generators in a multi-machine power system have been investigated. In everyday language, to“adapt”means to change a behavior to conform to new circumstances. Intuitively, an adaptive controler is a controller that can madify its behavior in response to changes in the dynamics of the proces and the disturbances. Because of the complex beha vior of adaptive systems it is necessary to consider them from several points of view. Theories of nonlinear systems, stability, system identification, recursive parameter estimation, optimal control and stochastic control all contribute towards the understanding of adaptive systems. Adaptive controller will contain i. Control law with adjustable parameters ii. Characterization of the closed-loop response iii. Design procedure iv. Parameter updating based on measurement v. Implementation of the control law These parts are somewhat different for different adaptive schemes but have many common features. On-line determination of process parameters is a key element in adaptive control. Some of them are recur sive least squares algoritm, projection algoritm. Adaptive control algoritms are. separeted two based on identification. Algorithms with Implicit Identifica tion: A very useful adaptive control technique is to specify a desired performance and measure the actual per formance against this performance. Such controllers are commonly known as the Model Reference Adaptive Control lers. A reference model representing the desired beha viour of the closed-loop system is driven by the same input as the controlled system. The regulator parameters are adjusted depending upon the error between the system output and the reference model output. Model reference adaptive algorithm may be describ ed as below : i. Choose a model representing the desired perfor mance xxix-ii. Compute the error between the outputs of the reference model and the actual system, iii. Based on the correlation. between this error and the system states, update the feedback gains of the system. The behaviour of the plant and of the system model be given by n th order state equations yp(t) = Ap yp(t) f Bp u(t) *m(t) = \ ^(t) +Bmu(t) where y is output state vector u is the input control vector A,B are matrices, and the subscripts p, m refer to plant and model respec tively. An error function can be formed by writing e(t) = ym(t) - yp(t) The aim of the control is to force this error function to zero. This can be done by choosing a Liapunov function. For a satisfactory application of this technique, the designer must have a good idea of the performance that the system is capable of achieving within the cont rol constraints. Otherwise the control limitation occurs and the system response will be substantially different to that of the reference model. Algorithms with Explicit Identification: Another class of adaptive controllers performs an explicit identification of system parameters or system transfer function. A number of identification algorithms have been developed using the discrete domain mathema tics. Sampled data design techniques are used to compute control in the following form: i. Select a sampling frequency, f, about ten times the normal frequency of oscillation, ii. Each sampling interval, T(=l/f), update the system parameters. Several techniques, in recur sive form, is usually used to identify the transfer function of the system, iii. Use the updated estimates to compute the cont roller parameters based on the control strategy chosen. These algorithms is usually known self -tuning regulator. -xxv -In the first chapter, it is described adaptive control and presented several types of adaptive control algo rithms. Furthermore/ the previous studies on adaptive contr ol techniques which is applied power systems have been explained in detail. In this thesis is used the Parameter adaptive cont rol technique proposed by Goodwin and Long. The used adaptive controller with an optimal predictor possesses the advantages of concise structure, a small number of measured output variables, a simple adaptive control algorithm. The design and operation of the controller do not require any information about the model parameters and persistent excitation of inputs of the turbogenerator are unnecessary. Also, the economical calculations of the adaptive control algorithm avoids problems arising with large sampling intervals. In the second chapter, it is presented identifica tion techniques and parameter adaptive control technique. The adaptive control scheme which is used, is illustrated in Figure 1. r~" -{Ste- ctont model ?^ s$- adapHve gains -«ag- adaptation, mechanism s-$£3G« optimqf predictor ?wca^-.~7 Figure 1. Schematic diagram of adaptive controller. In recent years, there has been considerable interest in application of adaptive control techniques to design controllers for turbogenerators with the objec tive of extending the operational margins of stability. -xv-In the third chapter, it is presented all of above calculations in detail. In the design of controllers, identification of turbogenerator models is based on autoregressive moving- average (ARMA) models. The models are obtained from the data of system inputs and local terminal state variables by using identification techniques. The identified ARMA models,, even though obtained from local measurements, include dynamic information from the other machines and the network. The ARMA model can be written in the following form* N M y(t+-l) = £ A±y(t-i) + S B.u(t-j) i=o j=o where y(Rm and u£Rm, A± (i=0,l,...,N) and B. ( j=0,1,...,M) are mxm dimensional real matrices, (m: number of input = number of output) A highly nonlinear differential equation is used to represent a turbo generator. Instead of this, it is used ARMA models. The outputs and inputs of each turbo generators are chosen as follows: yk(t) = [AVgk(t), Aa)k(t)]T uk(t) - [AVRk(t), AUGk(t)JT k=l,...,z where, z is the number of generating units and AVgk - Vgk-Vgok ^k^k^o AVRk - VRk'VRok AUGk = UGk-UGok Vgok and uo are respectively the set point of terminal voltage of each generation units and the steady state frequency of the power system.. V^o^ and UGok are the given values of the exciter input and the governor input of same generating units,, respectively. The degree of ARMA models is described by trial and error, so that N=2 and M=0 were selected to obtain a suitable minimum order model. In fourth chapter, it is presented the effect iveness co-ordinated control of exciter and governor, and the performance of the parameter adaptive controller in each generating units of the MMPS^was evaluted by compu ter simulation. The multimachine power system which is considered has five buses and four generating units. The -xvi-multi- machine power system is illustrated in Figure 2. Transient performance was assesed by simulating a 3-phase short circuit near 5th bus and then switched out trans mission line 1-5. Fault duration is 120 ms. Gl Trl S J, ->*i I Figure 3. Multi-input multi-output system configura tion. -XVXXIt has been shown in. the literature that the adaptive controller can be employed to control turbogenerators, with improvements in system stability. The results obtain ed by simulation studies and laboratory tests indicate that such controllers may provide substantial improvements in performance during various operating conditions, but almost all the results published so far relate. to a single machine connected to an infinite bus. Investigation of control strategies based on rep resentation of a single machine to infinite bus system is usually regarded as the first step in the study of turbo generator control. In these studies, interactions between generators are ignored. There is an infinite power source, which can maintain power system states around fixed levels. The stability of the generators only depends on its environmental. However, in a multimachine power sys tem (MMPS) without an infinite bus, system stability is decided by the relative movements between the machines. The behaviour of the controlled machine is affected not only by its own controller, but also the external dyna mics and controllers on the other machines. That is, the implementation of robust adaptive controllers in a multi- machine environment, control is exercised on each genera tor unit in a large interconnected dynamic system which is highly nonlinear. This thesis is concerned with an evaluation of a adaptive controller applied in a multi- machine power system. The evaluation involves the estab lishment of a detailed multimachine power system model, identification of turbogenerator models, design of the adaptive controller and assessment of performance in the multimachine power system. The mathematical model of multimachine power system consist of representations of turbogenerator units, trans mission networks, system load flow and equivalent loads, and also calculations for initialisation of multimachine power system, modification of nodal impedance matrix, and short circuit voltage and current. Each turbogenerator unit is represented by 14 th-order nonlinear differential equations which include the generator, excitation system, turbine and governor system. Synchronous generators are by seventh-order non linear differential equations including swing equations which are used to determine the angular diplacement between the machines of the power system during transient conditions. The representation is based on Park's refe rence frame, with components along the direct and quadra ture axes. In the rotating exciter system, the exciter itself is represented as a first-order linear system with a large time constant T as in this study. The saturation effect is ignored, and allowing a negative value for the field voltage ceiling limit. _ xviii-A three-stage turbine with reheat is employed to drive the synchronous generator. The steam is produced by a conventional coal-or oil-fired boiler, which is assumed to be a constant steam source. The inertia of steam flow in each stage of the turbine is described by a first-order transfer function, and the outputs from each are weighted according to their contributions to the total shaft torque, and summed. The reheater is also des cribed by a first-order transfer function. The steam flow is controlled by both main and interceptor values, with fast-acting electro-hydrolic governors. The value servo- mechanism is represented by a first-order transfer func tion, with limits on value travel and rates of movement. Since the transient behaviour of the electric net work is much faster than that of turbogenerators systems, the network may be described by algebraic equations. The network impedance equation is:. ». V = ZI. V is an nxl complex vector of nodal voltages measured with respect to the reference node. İ is an nxl complex vector, of injected nodal currents where n is number of node. Z is nodal impedance matrix. 'Z is an nxn complex matrix. It is known that since relative angular diplace- ments exist between individual machine rotors in a multi- machine power system, the positions of d-q coordinates with respect to the synchronously rotating d-q reference axes of individual machines are different. In order to carry out an integrated calculation of the power system, which includes turbogenerator units and the electric net work, common x-y coordinates are used as reference axes, into which the terminal voltages are current of various machines are transformed. In multimachine power system, the load flow prob lem results a set of nonlinear equations which can be solved by the Gauss-Siedel iterative method. The load flow calculations are achieved on complex x-y axes. The network variables obtained by the load flow are based on the reference frame, where the slack node, rather than the reference machine shaft, is chosen as the reference node. An operator of angular transformation is introduced to transfer the reference frame from the slack node to the reference machine shaft. The nodal impedance matrix can be modified to reflect changes in the network. These changes may be addition of elements, removal of elements, or changes in the impedances of elements. -xix-In the final chapter, it is presented the results which parameter adaptive control algorithm is simulated on MMPS which has multi-input multi-output generating units. r-XX-
Benzer Tezler
- Adaptive backstepping control based emegency scheme for improving transient stability of power systems with renewable energy sources
Yenilenebilir enerji kaynaklarıyla entegre güç sistemlerinde geçici kararlılığı kontrolu için uyarlanabilir geri adım kontrolü tabanlı acil durum düzeni
MOHAMMAD MOTALLEBZADEH
Yüksek Lisans
İngilizce
2022
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiElektrik Mühendisliği Ana Bilim Dalı
PROF. DR. VEYSEL MURAT İSTEMİHAN GENÇ
- Çok makineli güç sisteminde açısal kararlılık analizi ve kontrolör parametre optimizasyonu
Angular stability analysis and controller parameter optimization in multi-machine power system
SERDAR EKİNCİ
Doktora
Türkçe
2015
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiElektrik Mühendisliği Ana Bilim Dalı
PROF. DR. AYŞEN DEMİRÖREN
- Çok makineli güç sistemlerinde PSO algoritması yardımıyla geçici kararlılığın düzeltilmesi
Transient stability analysis of a multimachine power system using PSO algorithm
SERDAR EKİNCİ
Yüksek Lisans
Türkçe
2010
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiElektrik Mühendisliği Ana Bilim Dalı
PROF. DR. AYŞEN DEMİRÖREN
- Rotor açısı kararlılığının analizi ve optimizasyon tekniklerinin uygulanması
Rotor angle stability analysis ve optimization techniques
ZAKİRHUSSAİN FARHAD
Yüksek Lisans
Türkçe
2015
Elektrik ve Elektronik MühendisliğiGazi ÜniversitesiElektrik-Elektronik Mühendisliği Ana Bilim Dalı
ÖĞR. GÖR. SÜLEYMAN SUNGUR TEZCAN
YRD. DOÇ. DR. İBRAHİM EKE
- Sezgisel optimizasyon algoritmaları kullanılarak güç sistemi dengeleyicisinin optimal tasarımı
Optimal design of power system stabilizer using heuristic optimization algorithms
BURAK KAYA
Yüksek Lisans
Türkçe
2024
Elektrik ve Elektronik MühendisliğiKırıkkale ÜniversitesiElektrik-Elektronik Mühendisliği Ana Bilim Dalı
DOÇ. DR. İBRAHİM EKE