Research on Microgrid Scheduling Based on Improved Crow Search Algorithm

Figure 2

Diagram of the crow search algorithm. (a) During the search, an individual crow will start a random search in a sector centered on the current location. The sector area consists of the memory position, the current position, and its difference. Due to this unimodal search method, the crow’s flight activity lacks locomotion and diversity. As can be seen from Figure 2, the crow lacks its exploitation capability (the ability to search for candidate solutions) and it cannot efficiently search for other possible global optimum regions (e.g., g), which can easily fall into local optimum. (b) CSA is a single way to improve the diversity of the algorithm by generating probabilistic random solutions through its own experience, which leads to a high probability of entering a local optimum, premature maturity, and other drawbacks, especially for multidimensional optimization problems.