Methods Of Fatigue Distribution And Active Power Combined Optimization For Wind Farms

During the operation of wind turbines, duing to the influence of the randomness of the wind speed and dynamic wind turbine state, materials of wind turbine are bearing complicated alternating load. Under these loads, many microcracks initiate and grow inside materials, and materials performance gradually worsens. It is necessary to carry out maintenance before the wind tubine reaching fatigue damage threshold. This paper focuses on the unbalanced fatigue distribution over a wind farm and its optimization methods missing matching the active power demands of power grid. A series of control strategies are proposed to solve the problems. These methods are beneficial to improving the reliability of wind farm by avoiding excessive fatigue in wind turbines, thus reducing the fault probability and cutting down the temporary maintenance frequency of wind farm. As a result, the total maintenance cost can be reduced. The main contents of this paper are concluded as follows:1. The principle of fatigue distribution and active power coordination optimization is proposed. The basic idea is building a function that can model the relationship between fatigue damage and active power of wind turbines. The overall control strategy is making full use of the regulation margin of active power of wind turbines when power grid limits wind power generation. In this way, both the grid friendly and fatigue distribution of wind farm can be optimized.2. The control strategies and realization methods of wind turbine responding grid disturbance are presented. An improved finite state machine of pitch control with grid disturbance supporting is proposed. Moreover, a pitch control algorithm based on numerical calculation is developed. The two measures make it possible to regulate the output of wind turbines continuously and significantly, and meet the active power demands of power grid disturbance. The improved pitch control strategies are also designed to integrate with the existing pitch control system, they are suitable for both new type and existing type wind turbines. The simulation results have illustrated the effectiveness of the proposed methods.3. Based on analysing interaction mechanisms among wake effect, active power and fatigue of wind turbines, the optimization strategy of fatigue distribution and active power for offshore wind farms is proposed. Two steps are proposed to reduce the calculation workload of the control strategy. One is grouping the turbines by wake propagation path, the other is preseting a threshold value as optimization goal rather than minimizing the standard deviation of fatigue damage. A solution based on multi-agent technolog is proposed to realize the control strategy in a distributed model, and to make the control strategy could execute in existing computing devices and communication system of wind farm. The results of simulation of Horns Rev wind farm have illustrated that the proposed solution is available for offshore wind farm optimization.4. A combined control method of active power and fatigue distribution in complex terrain wind farm is proposed. Wind speed models of complex terrain are developed in two ways, the analytical way and the intelligent way, to describe the effect of wind speed caused by landforms and wake effects. Further, interaction mechanisms among wind speed, active power and fatigue of wind turbines in complex terrain are modeled and analysised. An unified optimization method for fatigue distribution and active power for complex terrain wind farms is proposed. Specially, an improved real-value coded genetic algorithm is put forward, that is suitable for both analytical and intelligent wind speed model. This algorithm can solve the optimization equations of wind farm real time on common industrial computer. The results of simulation have illustrated that the proposed solution is an availability way to optimize active power and fatigue distribution in complex terrain wind farm.5. A practical solution of fatigue distribution and active power coordination optimization for both onshore and offshore wind farms is presented. This method takes the idea that the reference active power of wind turbines should be in inverse proportion to the fatigue of them. With the characteristic of lite computing, this method is suitable for exsiting wind farm control system and do not need extra computing resources and platforms, it is beneficial for engineering implementation. Through simulation in Horns Rev offshore wind farm and a specific mountain wind farm, the generality and availability of the proposed method have been proved. It also demonstrates that this method is suit for any types of wind farms with high active power curtail raito.