| During 13th five-year plan and a period of time in the future, China’s energy development is faced with the arduous task of promoting energy production and consumption revolution, ensuring national energy strategy security, and realizing the transformation. Power system as an important area of energy transformation, low-carbon development and low-carbon transformation of the electricity sector has become critical. As China’s energy-consuming subject, the energy structure optimization of the power system lags far behind the developed countries, and to actively promote the low carbon transformation of China’s power system is the inevitable way to deal with the increasingly serious conflicts between economic development goals and issues including climate mitigation, environmental and energy security issues and so on.Firstly, this paper analyzes the transformation process and the key dynamic process by using multi-perspective model (MLP) based on the system transformation management theory and the process of new industry innovation development, researches the macro-picture of China power system, technological social paradigm and crack technology, constructs the system-policy-technical co-evolution model, analyzes deeply the interaction mechanism between system, policy, technology and market.Secondly, through the study of analyzing and choosing the learning curve model factors, the paper constructs two-factor learning curve model under the action of the cumulative amount of R & D and cumulative installed capacity, evaluates quantitatively the relationship between policy and technology innovation and applications clear, and researches the power technology learning rate and evolution trend of investment cost.Again, the paper builds the levelized power generation cost model via electric power technology financial parameters, investment parameters, operation and maintenance parameters, tax parameters, capacity parameters and so on, assesses the economy of major power technology comprehensively. In situations of maintaining the existing emission reduction measures, renewable energy would have the price advantage to compete with coal power after 2030, and by levying a high level of coal resource tax, carbon tax and environmental pollution tax, and photovoltaic electricity price subsidies, renewable energy electricity is below the coal price in 2020, and economy is obvious; basing on this, the paper proposes power system transition path in different stages of development.By setting the policy transition probability matrix, the paper establishes dynamic integrated resource strategic planning model based on Markov Decision Process, taking minimizing the total cost of the whole society as the goal and taking Electric power demand, power load, installed capacity, flexible power supply, pollutant emissions as the constrain conditions to clarify the impact of policy mix and implementation strength on power planning, and to clarify the time and scale of different power supply into the planning, uses particle swarm optimization (PSO) cultural algorithm to solve power planning path under different policy sets. In copy path of scene one, by continuing and developing the existing system structure and operation mode, and remaining technology development path dominated by coal power, coal as the main power supply ensures electricity demand. In deconstruction and reconstruction path of scene two, by breaking through crack technology blockade renewable power technology transforms into mainstream technology development path, and coal as the peaking plant provides support for renewable power, leading the transformation of the old society-technology paradigm. In technical alternatives and paradigm-rebuilding path of scene three and four, renewable energy power generation technology competing with the current mainstream of coal technology, coal will peak in 2020 and wind power and solar power, respectively in 2020 and 2030, will realize the large-scale development.Finally, the paper proposes the multiple attribute decision making system for electric power planning from the economy, environmental protection, adaptability three criteria based on the energy security concept in the new century, uses rough set theory to determine the index weight, and evaluates all aspects synthetically with using goal programming model for interval multiple attribute decision making, proposes the scene two as optimal path planning for power transformation.Through the above research work, the main innovation points in this paper are:Firstly, basing on the MLP model the paper constructs the system-policy-technical co-evolution model, researches the interaction mechanism between system, policy, technology and market, and builds the scientific management theory and analysis framework of power system transformation.Secondly, the paper innovates the methodology of power system low carbon transformation, and builds the "system-policy-path-planning" integration optimization decision support system with combination of qualitative and quantitative. Under the framework of multi-perspective analysis, the paper studies two-factor learning curve model, analyzes the quantitative relationship between the energy policy and technology innovation under the action of accumulated research and the cumulative installed, explores the dynamic effect mechanism of generating cost, electric power planning by the rule of technological learning, and forms a set of "macro picture-transition route optimization policy interaction-technology-scenario simulation" power transformation system optimization.Thirdly, the paper establishes dynamic integrated resource strategic planning model based on Markov Decision Process, through the design policy transition probability matrix, introduces the policy variable into power supply planning, and researches the implementation path planning of dynamic power under different policy in China in 2015-2050.Fourth, basing on the energy security of the new century, according to the economy, environmental protection, adaption, the paper builds multi-scale decision-making system of the electric power planning scenarios, establishes the interval multi-attribute decision-making goal programming model, makes comprehensive comparison and evaluation to the scheme of electric power planning scenarios. |