Data Mining Based Model Simplification and Optimization of An Electrical Power Generation System

ABSTARCT :

To assess the performance of electrification in an aircraft, multi-physics modeling becomes a good choice for the design of more-electric equipment. The high computational cost and huge design space of this complex model lead to difficulties in the optimal design of the electrical power system, thus, model simplification is mandatory. This paper first proposes a novel model simplification approach based on data mining, and the design of a small electrical power generation system is investigated to demonstrate it. According to the formulated multi-physics model of the system, The paper uses the optimal Latin Hypercube-based design of experiment to generate data for the analysis.Based on the generated data, a fusion algorithm integrating multiple feature selection methods is presented to facilitate the dimensionality reduction of the problem’s design space. And machine learning algorithms are applied to the surrogate model establishment, allowing the reduction of computational time.The investigation of various optimization routines with various multi-objective genetic algorithms shows that the proposed practices improve the system-level optimization efficiency with low computational complexity, ease of search, and high accuracy, which is competitive compared with state-of-arts.

EXISTING SYSTEM :

? The work in this thesis spans several multi-disciplinary research areas, all of which can be said to be related to the broad category of data mining. ? Data mining is indeed a broad subject, so it is important to provide some clarification. ? We introduce in this chapter some background on the topics this work covers, as well as background on what?conventional data mining tools can already provide. ? Existing off-the-shelf solutions can provide subsequent value to the work described in this thesis; these tools are likely to be useful from a practical standpoint but are not the focus of this research.

DISADVANTAGE :

? To overcome the inherent disadvantages of MLAs, the voting rule is implemented to synthesize different feature selection algorithms. ? In the modeling phase, several proper MLAs should be used and compared to select the best configuration, and then the simplified models are trained and verified. ? If the requirements cannot be satisfied, other MLAs or more samples are required to improve the performance. ? MOEAs tailored for high-dimensional problems are both of poor performance, compared with NSGA-II.

PROPOSED SYSTEM :

? This dissertation proposes a formulation of the emission-constrained GSP and its solution methodology involving generation maintenance scheduling, unit commitment, and CO2 cap-and-trade. ? The coordinated optimal maintenance scheduling and CO2 allowance bidding strategy is proposed to provide valuable information for GENCOs’ decision makings in both electricity and CO2 allowance markets. ? In this paper, the quadratic regression (QR), decision tree (DT), RF, rapid supervised method (RS), and gradient boosting (GB) are used to evaluate the inputs of the generator model.

ADVANTAGE :

? The advantages of the embedded technique are the same as the wrapper technique, but it is better than the wrapper technique in terms of computational complexity. ? Despite the promising performance of these technologies, most of them are shown to be of low efficiency or effectiveness in the system-level design of the electrified aircraft power system. ? such a technique to choose a subset of variables from the data, which provides us a way of reducing computation time, improving prediction performance, and a better understanding of the data.