哈尔滨市耕地集约利用水平预测及障碍因子诊断

【目的】如何高效利用土地资源、提高耕地集约度是当前科学研究的主要问题,但如何科学预测耕地集约度则一直是研究难点。【方法】基于哈尔滨市耕地集约利用现状,运用熵权、变异系数组合赋权法、BP神经网络、障碍因子诊断法等揭示哈尔滨市耕地集约利用水平发展趋势及耕地集约利用影响因素,为耕地集约利用发展方向及提高耕地集约利用度提供理论依据。【结果】结果表明:(1)1985~2014年哈尔滨市耕地集约利用水平平均值为0.346 2,处于相对集约状态;2015~2024年哈尔滨市耕地集约利用水平呈稳定上升态势,耕地集约利用度均值为0.475,处于高度集约状态;1985~2024年耕地集约利用水平以相对平稳态势发展,且并未出现较大波动,40年间哈尔滨市耕地集约利用平均值达0.409 6,处于较集约状态。(2)从准则层指标障碍度来看,耕地产出效率和投入强度对耕地集约利用水平的影响较大,并且呈现不断上升的趋势,耕地产出效率和投入强度年均增长速率为8.69%、7.82%,耕地利用程度和耕地利用可持续性年均下降速率为4.86%、5.73%,耕地集约利用水平的主要障碍因子是灌溉指数、单位耕地面积农业从业人员、农业产值比、单位播种面积粮食产量等。【结论】运用BP神经网络对哈尔滨市耕地集约利用水平进行科学预测,用障碍因子诊断法诊断耕地集约利用障碍因子,能为研究哈尔滨市耕地集约利用发展趋势提供理论支撑。

PREDICTION AND OBSTACLE FACTORS DIAGNOSIS OF INTENSIVE USE OF CULTIVATED LAND IN HARBIN CITY

How to make use of the land resources effectively and improve the cultivated land intensive degree is the main problem of the current scientific research.Based on the present situation of cultivated land intensive use in Harbin city, the paper revealed cultivated land intensive use trends and influencing factors by using the method of combination weighting in entropy and coefficient of variation, BP neural network, barrier factor diagnostic, and so on.It aimed to provide a theoretical principle for cultivated land intensive use.The results showed that: (1)the value of cultivated land intensive use was 0.3462 in 1985-2014 in Harbin city, which increased to 0.475 in 2015-2024 year and showed a high intensive state.From 1985-2024, the overall level of cultivated land intensive use was 0.4096 and showed a relative intensive state in the 40 years.(2)The impact of cultivated land output efficiency and investment intensity on intensive use of arable land become large and presented a rising trend.The land production efficiency and input intensity annually increased 8.69% and 7.82%, while the land use degree and land use sustainability decreased 4.86% and 5.73%.the main obstacle factors of cultivated land intensive utilization level were irrigation index, farmers in per unit of cultivated land area, agricultural output ratio, grain yield per unit of sown area, and so on.It concluded that the method of BP neural network could provide the theoretical principle for predicting the trend of cultivated land intensive use in Harbin city.