Garden is an important support for regional economic development, but also an important support for regional ecological environmental protection, the rational allocation of water resources in the garden is one of the effective ways to solve the problem of water shortage. This paper takes the Internet of things, digital twin as the technical basis, uses the multi-objective optimisation algorithm to construct the water resource management model of the garden area, and uses the artificial fish swarm algorithm to solve the model. By constructing a digital twin irrigation district water resources scheduling management platform, the water resources elements of the garden area are comprehensively monitored and sensed, and the intelligent simulation of the water resources allocation management process and decision-making scheme evaluation and optimisation are achieved, so as to enhance the intelligent and refined management level of water resources scheduling of the garden area, and comprehensively realise the saving and intensive use of water resources. Taking X garden area as a research case, the water resources management model finally derives the optimal water resources allocation scheme under 50%, 75% and 90% in each planning year, which provides support for the efficient use of water resources in X garden area.