A new practical method is proposed which gives a quasi-optimal solution of the dynamic load dispatching problem formulated by a multi-constraint multi-objective optimization problem. Here, the multiple constraints are rate reserve constraints and power flow ones, and the multiple objectives are fuel cost and CO₂ emission.
The proposed method is an integration of the following three techniques: (1) “Group Dispatch Scheme”, which is developed by the authors, is used to satisfy a violated consraint easily. This scheme consists of three steps: The first step is to classify all generators into two groups. One is the group which should be more loaded to satisfy the violated constraint, and the other is the should-be less loaded group. The second step is to add some load to the former group, and to subtract the same load from the latter. The third step is to dispatch each group's load to each group's generators respectively. (2) The dynamic load dispatching algorithm is used which provides, in practical time, a quasi-optimal generation trajectories. (3) Te conventional weighting parametric method is used to obtain a Pareto optimal solution of multi-objective problems. The effectiveness of the method is clarified by a computer simulation on the actual power system of Kansai Electric Power Company.

View full abstract

Recently, power system becomes more complex. Generation plants are located far from the load center. Moreover, load shape becomes more sharp. Difference between peak and off-peak demand is spreaded more and more.
On the other hand, introduction of economical dispersed generators near load center and time of use pricing are considered. If many dispersed generation plants will be operated and time of use pricing will be introduced, above problem may be cleared.
In this paper, the effect of introduction of dispersed generation and time of use pricing as new pricing strategy at one area is studied. As a result, since the introduction of too many dispersed generation plants makes load factor lower, total surplus in the area decreases. However, in case if time of use pricing is introduced, the load factor becomes higher. Total surplus in the area can be increased.

View full abstract

This paper proposes a new methodology for planning of power supply program.
Firstly we suggest that the power supply program should be planned not monthly but weekly and be evaluated by the loss of load probability (LOLP), which assesses the degree of balance of power demand and supply at the time of weekly peak load, for each week about 52 weeks a year. Because we don't think that the monthly power supply program is adequate since the daily or weekly peak load changes remark-ably in such transient season as the period from spring to summer or from summer to fall. In case of large power system the calculation method by the convolution for assessing LOLP is not practical due to the complicated calculation procedure and the long calculation time.
So we secondarily propose to use the Gram-Charlier series expansion for a probability distribution function to calculate the LOLP. This series has a high degree of approximation for a number of units and very short calculation time.

View full abstract

A new method has been developed for thermal and pumped-storage hydro unit commitment, which can prepare the optimized schedules with higher security and lower operation cost. The method of thermal unit commitment consists of three steps. (1)The optimal priority order of each unit is prepared based on the ratio of fuel to power cost as a function of incremental fuel cost λ. (2)The solution candidates are selected by fixing higher priority units as must-on and lower priority as must-off, and by searching middle priority units as “search window” to satisfy the constraints depending on the interval such as spinning reserve and load flow. (3)Based on the above-selected candidates, the optimal schedule is determined by dynamic programming, which satisfies the constraints extended over the whole time intervals such as crew constraints. The hydro schedule is decided as a substitute for the peak thermal units of lower efficiency so as to further reduce the total operation cost.
The effectiveness of present method has been demonstrated by application to test problems for a real power system with 53 thermal and 3 pumped-storage hydro units, in which the present method can reduce the operation costs by 0.20%-0.43% compared to the conventional priority order method.

View full abstract

This paper presents a fuzzy-based method for determining a flexible generator maintenance scheduling by means of subjective relaxation of constraints imposed on maintenance scheduling problem. The constraints are divided into hard (crisp) constraint set and soft (fuzzy) constraint set in terms of reflecting conditions which surround power systems. The problem is formulated as a fuzzy mathematical programming problem and solved with the fuzzy branch and bound method using Bellman-Zadeh maximizing decision. The proposed approach provides not only new flexible concept of planning problems in power systems but also natural expansion of conventional approaches based on crisp set theory. The effectiveness and feasibility of the proposed approach are demonstrated on two typical power system models which consist of 15 generators and 60 generators, respectively.

View full abstract

Many papers have been reported on the daily unit commitment problem so far. But, as the week end load pattern is different from that of week-day, daily unit commitment method is not directly applicable to the weekly unit commitment problem. In this paper, an efficient method for the weekly thermal unit commitment schedule with various constraints is presented. As weekly unit commitment is a complicated and time consuming problem, a simple and efficient technique is required for the practical applications. Using an appropriate initial operating schedule by Dynamic Programming, an iterative approach which simply needs a few iterations is proposed. A multi-stage screening algorithm is developed to deal with five constraints, that is, unit's minimum up time, minimum down time, loading rate limit, fuel consumption and spinning reserve requirements.
Numerical examples obtained by a personal computer show the effectiveness of the proposed method in terms of computing burden and accuracy. The propriety of the obtained weekly unit commitment schedule is also discussed.

View full abstract

This paper presents an application of parallel genetic algorithms (PGA) to the optimal long-range genera-tion expansion planning. The problem can be formulated as a combinatorial optimization problem that determines the order of introducing generation units at each interval of the year. The proposed method considers introduced power limits of each technology, maximum loads at each interval, and load duration curves at each interval. Appropriate string representation for the problem is presented. Binary and decimal coding, and three selection methods are compared. The method is developed on transputer that is one of the parallel processors. The feasibility of the proposed method is demonstrated using a typical expansion problem with four technologies and five intervals and compared with the conventional dynamic programming and simple genetic algorithm with promising results.

View full abstract

This paper discusses optimal planning problems for dispersed generating sources so as to give a guideline of their introduction to power systems. First, a recursive method is developed from the viewpoint of supplying electric power only, in which solutions for one unit are superposed. The computational loads are decreased by selecting alternatives of busses based on the so-called penalty factor. However, for effective utilization of dispersed generating sources, it is indispensable to consider their thermal merits in addition to electric power. After applying the Khun-Tucker theory to the optimal dispatching problem including boiler fuel costs, thermal values of the dispersed generating units are evaluated quantitatively. Finally, the above-mentioned technique for the optimal planning is also effective basically and is extended into this case. The validity of the proposed algorithms are demonstrated using a real-scale model system.

View full abstract

In this paper, a fuzzy controller for a static Var compensator (SVC) to enhance the damping of power systems is proposed. The proposed fuzzy controller can consider the stability of power systems because the control rules of the fuzzy controller are constructed to satisfy the stability of power systems based on sliding mode control. Therefore, the control efficiencies are improved sufficiently; besides, the fuzzy rules can be composed systematically. The proposed controller has the robustness in terms of parameter variations and non-linearity. The validity of proposed control method is examined by computer simulations modeled on a single-machine infinite bus system.

View full abstract

During recent years, it has become remarkably important to protect from the voltage instability of power systems according to the tendency of hevily loading and the concentration of air conditioning loads on certain hours. The many studies have advanced elucidations on the problems and the countermeasures are filling up. However, the load representations used in those studies of the phenomena are based on power values, i. e., P and Q, therefore these analytic models are insufficient to express the load responses strictly, and feasible inverstigation regions are limitted.
In this paper, a load model expressing load admittance responses by a first order delay differential equation is introduced and studied to apply this model combining with power network solutions for analyses of the power system voltage stability, then this load model is shown to work reliably and effectively. Using this model, it is possible to investigate more precisely and widely than before the power system voltage stability phenomena such as one on arbitrary characteristic loads, in lower voltage ranges and on propagations of instability through the networks.

View full abstract

Recently, as part of preventive maintenance for equipment, there have been greatly increasing needs to diagnose the insulation of cables under hot-line conduction. In a background, there is a large number of equipment which was constructed around the high-growth period in the 1965s and have become obsolete, in addition to recent circumstances which make it difficult to obtain longer service interruption time.
In this paper, therefore, the results of study on the method measuring the live-wire insulation resistance by superimposing DC 50 V to a high-voltage busbar through GPT and forming a bridge with the resistane of a cable insulation taken as a side, are discussed. It was clarified that this live-wire insulation diagnostic method can detect the degradation of a cable insulation resistance from the high level and is therefore suitable for so-called trend monitoring as well.

View full abstract

With rapid growth of the number of literatures on nonlinear systems, nonlinearities existing in the systems under study have drawn much attention. Studies on the effects of nonlinearities in power systems have been becoming an increasingly important part of researches for the investigation of system stability. It is probable that hidden phenomena may unexpectedly emerge due tc the nonlinearities that are involved in load flow equations, generator swing equations and characteristics of control equipments and loads, etc.
This paper presents a new Catastrophe Theory application to nonlinear power systems. Making use of the concept of Duffing's equation, we show that a Catastrophe Theory analogy can be used to interpret unstable phenomena caused by system nonlinearities from the viewpoint of oscillations. When considering the system nonlinearities, due to ill-combination of system parameters and periodic disturbances, there may exist the characteristic “jumps” in system states corresponding to slow (quasi-dynamic) changes of the frequencies of periodic disturbances. With this Catastrophe Theory approach, a system bifur-cation set can be identified to assess the unstable phenomena of power systems.

View full abstract

In this paper a numerical model of the cogeneration system is proposed for daily operation scheduling. Especially this paper describes a numerical model considered both the steam to electricity ratio and the thermal storage tank for more reasonable operation of cogeneration system. Minimum daily operation costs of topping cycle and bottoming cycle are calculated by using the proposed model. Simulation results of four cases are analyzed and compared each other. By these simulations which are similar to real operation circumstances, the validity of the proposed model is verified.

View full abstract

The liquid-helium-cooled metallic superconducting cable can transmit a few giga-watts of electric power through a restricted small space. Although no resistive loss produces in it, a large amount of additional electric power is required to operate the refrigerators to keep it in the cryogenic temperature. This power must be included in the transmission loss of the superconducting cable. This is regarded as one of its large demerits. However, the superconducting cable would be operated in the established power system. Thus, its transmission loss should be evaluated together with the whole transmission loss of the system. In this paper, two types of a future metropolitan electric power system are supposed. In a “hybrid system”, the superconducting cables transmit a large portion of electric power generated in a satellite power plant to a demand center directly, while, in a “conventional system”, the future demand in this metropolitan area is supplied only by enhanced conventional transmission lines. It is verified through numerical simulations that the total transmission loss energy in kilowatt-hour is smaller in the “hybrid system” than that in the “conventional system” throughout a year. The total transmission loss in kilowatt is also smaller at a peak load period. These beneficial state would be sustained in a future long period when the electric demand would continue to increase.

View full abstract