This paper proposes a backward search approach to incorporate excess stream inflows, in stochastic dynamic programming (SDP) based reservoir scheduling, of hydropower plants. The aim is to reduce the amount of spill d...
详细信息
ISBN:
(纸本)9781424438105
This paper proposes a backward search approach to incorporate excess stream inflows, in stochastic dynamic programming (SDP) based reservoir scheduling, of hydropower plants. The aim is to reduce the amount of spill during operation of the reservoir. Energy maximization type stochastic dynamic programming (SDP) model, with inflow process assumption considered as independent, is developed for a hydropower plant located in Japan. Reservoir operating policy consists of target storage levels at the end of a period, for each combination of beginning of the period storage levels and possible average inflow states during the period. The backward search approach has been incorporated in the SDP model, during the simulation, in order to identify the annual reservoir operation curve, with reduced spill. The storage guide curves identified with incorporation of backward search approach are compared with that from the normal SDP model. It has been found that the proposed approach serves better from the view point of spill reduction.
The largest difference between cognitive radar and other adaptive radar is the adaptivity of transmitter in cognitive radar. How to optimally decide or select the radar waveform for next transmission based on the obse...
详细信息
ISBN:
(纸本)9781424442461
The largest difference between cognitive radar and other adaptive radar is the adaptivity of transmitter in cognitive radar. How to optimally decide or select the radar waveform for next transmission based on the observation of past radar returns is one of the important issues. In this paper, with the stochastic dynamic programming model of waveform selection, we use the method of temporal difference learning to solve this problem and realize the adaptivity of waveform selection. The simulation results show that the uncertainty of state estimation using temporal difference learning is less than that using fixed waveform.
This paper presents a predictive control approach for long-term generation scheduling of hydro-thermal power systems. The approach is based on an open-loop feedback control scheme that uses a neural fuzzy network fore...
详细信息
ISBN:
(纸本)9781424438105
This paper presents a predictive control approach for long-term generation scheduling of hydro-thermal power systems. The approach is based on an open-loop feedback control scheme that uses a neural fuzzy network forecasting model, for handling the stochastic nature of inflows, and a deterministic nonlinear optimization model, to determine the discharge decisions to be implemented. As a consequence, inflow correlations on time can be represented by nonlinear relationships, and hydropower generation and thermal fuel cost can be handled by nonlinear functions, allowing a more precise modeling of the problem. A simulation model was also developed for performance assessment of the proposed approach. A comparison with the classical stochastic dynamic programming approach, in the case of single reservoir systems, revealed that the latter and the proposed approach perform similarly. The approach was also applied to a multi-reservoir system composed of 19 hydro plants and 10 reservoirs corresponding to a major cascade of the Brazilian power system. The results show that the proposed approach performs as well as in the single reservoir case study.
The quality of environmental decisions are gauged according to the management objectives of a conservation project. Management objectives are generally about maximising some quantifiable measure of system benefit, for...
详细信息
ISBN:
(纸本)9780975840078
The quality of environmental decisions are gauged according to the management objectives of a conservation project. Management objectives are generally about maximising some quantifiable measure of system benefit, for instance population growth rate. They can also be defined in terms of learning about the system in question, in such a case actions would be chosen that maximise knowledge gain, for instance in experimental management sites. Learning about a system can also take place when managing practically. The adaptive management framework (Walters 1986) formally acknowledges this fact by evaluating learning in terms of how it will improve management of the system and therefore future system benefit. This is taken into account when ranking actions using stochastic dynamic programming (SDP). However, the benefits of any management action lie on a spectrum from pure system benefit, when there is nothing to be learned about the system, to pure knowledge gain. The current adaptive management framework does not permit management objectives to evaluate actions over the full range of this spectrum. By evaluating knowledge gain in units distinct to future system benefit this whole spectrum of management objectives can be unlocked. This paper outlines six decision making policies that differ across the spectrum of pure system benefit through to pure learning. The extensions to adaptive management presented allow specification of the relative importance of learning compared to system benefit in management objectives. Such an extension means practitioners can be more specific in the construction of conservation project objectives and be able to create policies for experimental management sites in the same framework as practical management sites.
For metapopulation management problems with small state spaces, it is typically possible to model the problem as a Markov decision process (MDP), and find an optimal control policy using stochastic dynamic programming...
详细信息
ISBN:
(纸本)9780975840078
For metapopulation management problems with small state spaces, it is typically possible to model the problem as a Markov decision process (MDP), and find an optimal control policy using stochastic dynamic programming (SDP). SDP is an iterative procedure that seeks to optimise a value function at each timestep by trying each of the actions defined in the MDP. Although SDP gives the optimal solution to conservation management questions in a stochastic world, its applicability has always been limited by the so-called curse of dimensionality. The curse of dimensionality is the problem that adding new state variables inevitably results in much larger (often exponential) increases in the size of the state space, which can make solving superficially small problems impossible. A large state space makes optimal SDP solutions computationally expensive to compute because optimal SDP techniques require the value function to be updated for the entire state space for every time step. The high computational requirements of large SDP problems means that only simple population management problems can be analysed. In this paper we present an application of the on-line sparse sampling algorithm proposed by Kearns, Mansour & Ng (2002), which can be used to approximate the optimal solution of a MDP for a given starting state. The algorithm is particularly attractive for problems with large state spaces as it has a running time that is independent of the size of the state space of the problem. We apply the algorithm of Kearns et al. (2002) to a hypothetical fish metapopulation where we define the management objective to be to maximise the number of occupied patches during the management horizon. Our model has multiple management options to combat the threats of water abstraction and waterhole sedimentation. We compare the performance of the optimal solution to the results of the on-line sparse sampling algorithm for a simple 3-waterhole case. We find that the approximation algorithm out
Wind generation is currently undergoing the fastest rate of growth of any form of electricity generation in the world. Integration of wind power into systems could be problematic, however, due to availability and vari...
详细信息
ISBN:
(纸本)9781424442409
Wind generation is currently undergoing the fastest rate of growth of any form of electricity generation in the world. Integration of wind power into systems could be problematic, however, due to availability and variability of wind supply. As a proven storage technology, pump storage may provide/absorb additional capacity so as to hedge against adverse situations. A scheduling software is needed to coordinate wind generation and pump storage units, given existing price and wind forecasting modules. This paper first implements a "Collocation Method" to solve a dispatching problem of wind and pump storage units, given the forecasting of future price and wind generation. Second, given that a pumping decision is discrete while generation dispatching is continuous, this paper uses a two-step strategy to solve a mixed-integer problem. Transition constraints are modeled based on the property of collocation method. Third, this paper uses real wind generation and real-time energy price data to simulate scheduling activities by the proposed model. A comparison of expected profits between wind only and wind plus pump storage unit is given.
This paper deals with a stochastic optimal control problem for a class of failure prone buffered multi parts flow-shops manufacturing system. The flow shop under consideration requires setup time and cost in order to ...
详细信息
This paper deals with a stochastic optimal control problem for a class of failure prone buffered multi parts flow-shops manufacturing system. The flow shop under consideration requires setup time and cost in order to switch the production from a part type to another. Our objective is to find the production plan and sequence of setups that minimize the cost function which penalizes inventory/backlog and setup costs. A continuous dynamicprogramming formulation of the problem is presented and solved numerically for a two buffered serial machines two parts case. It will be shown that the obtained policy is a combination of a KANBAN/CONWIP and a modified hedging corridor policy.
This paper examines an economy with a large number of industries, each producing a different good. Technological change follows a Poisson process where firms improve their productivity through investment in R&D. T...
详细信息
This paper examines an economy with a large number of industries, each producing a different good. Technological change follows a Poisson process where firms improve their productivity through investment in R&D. The less there are firms in the economy or the more they can coordinate their actions, the higher their profits. Labor is used in production or R&D. All workers are unionized and their wages depend on relative union bargaining power. If this power is high enough, then there is involuntary unemployment. Both workers and firms lobby the central planner of the economy which affects firms’ and unions’ market power. The main findings of the paper can be summarized the follows. Unions’ and firms’ market power decreases the level of income at each moment of time. On the other hand, the greater the firm's share of value added or the higher union wages, more incentives the firm has to increase the productivity of labor through R&D. In this respect, there can be an optimal amount of unions’ and firms’ market power. Concerning the regulation of relative union bargaining power, the central planner can increase its welfare either (a) by increasing the level of income or (b) by speeding up economic growth. If (a) is more effective than (b), then the central planner eliminates union power altogether to have full employment. On the other hand, if (b) is more effective than (a), then the central planner supports labor unions to promote cost-escaping R&D.
The largest difference between cognitive radar and other adaptive radar is the adaptivity of transmitter in cognitive radar. How to optimally decide or select the radar waveform for next transmission based on the obse...
详细信息
The largest difference between cognitive radar and other adaptive radar is the adaptivity of transmitter in cognitive radar. How to optimally decide or select the radar waveform for next transmission based on the observation of past radar returns is one of the important issues. In this paper,with the stochastic dynamic programming model of waveform selection,we use the method of temporal difference learning to solve this problem and realize the adaptivity of waveform selection. The simulation results show that the uncertainty of state estimation using temporal difference learning is less than that using fixed waveform.
By the theory of stochasticdynamic program- ming,we provide the methods for deriving the optimal *** this paper,we make two models in dynamic state process to maximize the expected utility of the agent and then obtai...
详细信息
By the theory of stochasticdynamic program- ming,we provide the methods for deriving the optimal *** this paper,we make two models in dynamic state process to maximize the expected utility of the agent and then obtain the famous Hamilton-Jacobi-Bellman equation. Furthermore,we derive explicit form solution and closed- form solution of the optimal equations for given utility functions.
暂无评论