| Since the industrial revolution,a large number of greenhouse gas emissions have led to an increasingly severe climate situation.The development of low-carbon economy has become a strategic consensus of all countries in the world to deal with climate problem.In order to realize the low-carbon economy,it is particularly critical in the construction field,where global energy consumption and greenhouse gas emissions account for nearly one third.The building integrated energy system with the advantages of multi energy complementarity and energy cascade utilization is an important solution for the development of low-carbon buildings.On the other hand,with the application of emerging technologies such as big data and deep reinforcement learning in the construction field,the related concepts of intelligent buildings have developed rapidly,and the research on energy adaptive optimization of intelligent buildings in uncertain environments is also facing new development opportunities.However,due to the bilateral uncertainty of supply and demand of intelligent buildings and a large number of system random variables,the research still faces great challenges.Therefore,this thesis focuses on the energy self-adaptive optimization of intelligent buildings in uncertain environment.The main work is as follows:(1)A energy optimization method of intelligent building considering the uncertainty of both supply and demand is proposed.Considering the different sources and characteristics of the load demand of intelligent buildings and the output uncertainty of renewable energy,interval numbers and opportunity constraints are used to represent them respectively.An interval optimization model with opportunity constraints is established.After it is converted into a standard interval linear programming model,the enhanced interval linear programming algorithm is used to solve it.The experimental results show that the proposed algorithm can provide a more flexible scheduling scheme than the optimization method without considering the uncertainty characteristics.(2)A joint optimization method for energy and thermal comfort of intelligent buildings under complex uncertain environment is proposed.On the basis of both supply and demand uncertainties,considering the thermal comfort needs of users,outdoor temperature,building thermal dynamic model parameters and other system random variables,a Markov decision process is constructed and solved by an improved depth deterministic strategy gradient algorithm.The effectiveness and feasibility of the proposed algorithm are verified by experimental analysis.The experimental results show that the training efficiency and stability of the proposed algorithm are improved compared with the original algorithm,which can reduce the total operation cost of intelligent buildings by 4.8% and the total temperature deviation by 56.1%,and has the ability to adapt to uncertain environments. |
