| 摘要: | 人類活動產生大量的溫室氣體，加劇了溫室效應的影響，使全球的平均溫度持續上升，而全球暖化造成的各種影響，使生活環境產生極大的變化。世界各國為了讓全球的平均溫度不再持續增加，便各自提出了眾多解決方案，而利用可再生能源取代化石燃料，能夠大幅降低溫室氣體的排放量，因此世界各國廣設太陽能板。太陽能板的發電效率會受到許多因素影響，其中環境是影響發電效率的主要原因之一，環境中的灰塵或動物糞便等髒污皆會在太陽能板上產生陰影，故太陽能板的清潔尤為重要。在傳統上是以人力進行太陽能板的清潔，但太陽能板的設置區域為屋頂或是大規模的太陽能發電廠，利用人力清潔不但費時，還有可能發生意外。若使用太陽能板清潔機器人並配合無人機進行運送，則能夠避免意外的發生，還能夠滿足分散式的太陽能板清潔需求。
由於市場上存在多種性能不同的太陽能板清潔機器人，其清潔速度、電量和可適用的最大傾斜角度等性能差異，將影響完成清潔太陽能板所需的總時間。因此，本研究的目標是在滿足所有清潔需求的前提下，選擇合適的太陽能板清潔機器人配置，以最小化總完工時間。本研究考慮了五種不同性能的太陽能板清潔機器人，並使用隨機生成的太陽能板相關資料（區域編號、面積大小、傾斜角度）進行分析。在太陽能板清潔機器人配置問題中，為了符合實際應用情境，而假設每個區域僅使用一種太陽能板清潔機器人的限制，並採用基因演算法來進行優化。以染色體來表示太陽能板清潔機器人的配置，並根據染色體的資訊對相同編號下的太陽能板清潔機器人進行派工。透過電腦實驗得知基因演算法能夠處理太陽能板清潔機器人的性能差異、區域的面積大小和傾斜角度等限制與變數。此研究能夠為太陽能板清潔領域提供實用的解決方案，並在選擇合適的清潔機器人配置時做出合理的決策。這將有助於提升清潔作業的效率、減少清潔所需的時間和人力成本。;Human activities emit a considerable quantity of greenhouse gases, intensifying the greenhouse effect and causing the global average temperature to increase further. The numerous effects of global warming have drastically altered the living environment. Countries throughout the world have proposed a variety of remedies to keep the global average temperature from rising further, and the use of renewable energy to replace fossil fuels can significantly decrease greenhouse gas emissions. As a result, countries all around the world have widely installed solar panels. Many elements will influence the power generation efficiency of solar panels, with the environment being one of the most important. Because dirt in the surroundings, such as dust or animal dung, casts shadows on solar panels, cleaning solar panels is critical. Traditionally, solar panels are cleaned by manpower, but the installation area of solar panels is a roof or a large-scale solar power plant. Using manpower to clean is not only time-consuming, but also may cause accidents. If a solar panel cleaning robot is used and transported with a drone, accidents can be avoided, and decentralized solar panel cleaning needs can also be met.
Because there are a number of solar panel cleaning robots on the market, differences in performance such as cleaning speed, power consumption, and appropriate maximum tilt angle will impact the overall time necessary to finish solar panel cleaning. Therefore, the purpose of this research is to choose a proper solar panel cleaning robot configuration that minimizes overall completion time while achieving all cleaning requirements. In this study, five solar panel cleaning robots with different performances were considered and analyzed using randomly generated solar panel data. In the solar panel cleaning robot configuration problem, in order to meet the actual application situation, it is assumed that each area only uses one solar panel cleaning robot, and the genetic algorithm is used for optimization. The configuration of the solar panel cleaning robot is represented by the chromosome, and the solar panel cleaning robot under the same number is dispatched according to the information of the chromosome. Through computer experiments, it is known that the genetic algorithm can handle the performance differences of solar panel cleaning robots, the size of the area and the angle of inclination, and other constraints and variables. This will help improve the efficiency of cleaning operations, reducing the time and labor costs required for cleaning. |
