Agricultural machinery selection and scheduling of farm operations
MetadataShow full item record
The procedure of calculating annual machinery ownership costs from the discounted cash flows of the mortgaged capital cost, the repair and insurance charges and the resale income is extended to include the effect of loan rate and loan period on interest charges, the effect of capital allowances taking account of the actual balancing charges at the end of the period of ownership, and the effect of tax relief on the interest charges, repair costs and insurance premiums. The concept of marginal holding cost is applied to determine the optimum ownership period.The selection of tractor- plough combinations is based on the prediction of soil characteristics such as moisture content, strength, and workability, all of which influencing the assessment of plough draught and tractor power. A number of filters are used to select the appropriate and realistic tractor /implement combinations with different sizes of fully mounted plough depending on the draught, and the speed of each selected gear of the tractor. For each acceptable combina- tion of tractor and fully mounted plough determined, the costing routine is used to calculate the annual costs.The branch and bound algorithm is suitable for mixed integer solutions to the farm machinery selection problem. Machinery sets are selected simultaneously with the chosen cropping pattern on a given land area. Machinery sets are matched correctly to the tractor sizes. Four sizes of tractor are available (45 kW, 61 kW, 74 kW and 94 kW,. Field operations take place in discrete time periods during which available work days are predicted from soil type and weather records for the specific site. Cereal and root crops are distinguished by optimum sowing and harvesting date. Discrete time periods are defined in relation to these optimal dates and give rise to overlapping operations for different crops. The calculation of probability levels for available work days when operations are subject to different criteria is discussed. A single arbitrary value of 75% probability for available work days is adopted in the linear programming model for the main part of the study.Two stage processes are used to simulate available work days in each time period. The patterns generated converge on the relative frequency pattern laid down by the generating process. The range of experience is wider than that contained in the short series of 24 years historical data. The simulation model generates results suitable for stochastic dominance ranking.In a simulation experiment on a 250 ha arable farm cropping cereals and potatoes, alternative solutions are obtained by integer linear programming, the solutions being ranked according to gross revenue. Annual costs of operating farm machinery are derived from a separate costing algorithm based on the annual hours of use which are determined by the size of the task and not by the sequence of work days. After deducting the annual costs of machinery operation, the cumulative net revenue curves cross and second order stochastic dominance ranking is used to identify the optimum (maximum profit) solution.The current study demonstrates the viability of the analytical procedures but further work is now required to reduce the computing time involved for the complete machinery selection procedure. Meanwhile, a commercial software package is prepared on the calculation of annual machinery ownership costs.