| Farm machinery system is composed of a number of factors such as crop, soil and farm machinery etc., the function of which is to perform field operations relating to crop production in specific weather and soil conditions. Research show that field operations by farm machinery may account to 30-50% of total cost for crop production in China, making the task of optimization of farm machinery system significant both in theory and practice.Though a number of researches have been conducted on the farm machinery systems, some key problems still remain unresolved. One of which is how to correctly select suitable farm machineries, as most studies were limited to theories and lacking enough operational approaches to provide guidance to farmers. The second problem is the determination of the depreciation cost, of which the linear model of depreciation cost were generally adopted, deviating largely from the real situation for most types of farm machinery. The third is the repair and the maintenance cost, here attention were not enough paid to the application of machinery information, such as machinery type, age, cumulated working hours and field working procedure etc. The forth is the machinery field workable probability, which were occasionally judged empirically other than enough consideration on the local weather and soil conditions. Another problem is the lack of operations from timeliness cost of key operations, due to the unavailable experience function of timeliness cost for most regions, in most cases key field operations were limited to perform in fixed period. The last problem is the ignorance of the reliability degree of farm machinery upon optimization.The customer satisfaction degrees of farm machinery were firstly surveyed and analyzed, yielding results that can be applied as a basis for farm machinery selection. After that a number of sub-models based on real farm data were studied, which include value index model for farm machinery and those relating to machinery field workable probability and field operation cost forecast. Again the provision of these sub-models facilitated the optimization of a renovated model. For the purpose of evaluation of the new models, field tests were conducted on typical rice production systems with different features and evaluations were performed on each rice production regions in China.In order to analyze customer satisfaction of different farm machinery products, a questionnaire was designed in accordance to Likert scale. And then totally 9982 questionnaires were collected from Jiangsu Province, sources of which came from managers, extension workers and drivers familiar or expertise in this area. The investigation covered five types and 150 specific products made from 50 different manufacturers. The five types of farm machinery include tractor, straw-return rotary tiller, planter, combine harvester, and rice transplanter. Their grades, distribution and ite correlations to customer satisfaction were analyzed in SAS software. Results show that customer satisfactions on the five types of farm machinery products were ranked, from high to low, as tractor, straw-return rotary tiller, planter, combine, and rice transplanter, respectively. The corresponding rates of products number related to the customer satisfaction surpassed the "general satisfaction" to 91.23%,100%,94.12%,87.10%and 81.82%, respectively. And those higher than the "relative satisfaction" were 84.12%, 91.18%,88.24%,41.94% and 72.73%, respectively. The detailed analysis yielded the ranks of customer satisfactory for each manufacturer and each 13 different series products, showing that the customer satisfaction correlation for combines manufactured by the same manufacturer was 0.9061. These achievements have already been successfully used to manage farm machinery in Jiangsu province.To develop new residue value index models for medium tractors,128 tractor auction sales data in the past 15 years in Shanghai State Farm were collected. From them sub 116 samples were randomly selected as regression samples, and the remaining 12 data were designated as testing data. Different Box-Cox transformations and regression methods were carried out in SAS, showing that the double square root model was best one among the 6 proposed explanatory models for tractor age and average yearly working hours. Tractor age was the main factor for determining tractor residue value index. The test on the real data also show that the forecasting errors of the model were within±10%for 8 of the 12 tested samples, the maximum one is-17.15%, and that of ASABE model were within±10%only for 3 of the 12 tested samples, the maximum one is 47.59%. This proved that such models can supply satisfactory forecasting results in similar situations.Field workable probability of farm machinery was predicted with weather data from the last 15 years and then it was linked to models developed. The results show that evaporation models of the first half year and the second half year were satisfactory (adjusted sum square were 0.7985 and 0.7167 respectively). Both average daily temperature and daily sunshine hours affect evaporation dramatically, with the evaporation in the first half year slighter higher under the same temperature and sunshine hours. The real data test also show the forecast errors of the models, ranging in±20% for 9 of the 12 tested samples. Markvo model was a reasonable tool for the forecast of monthly rainfall, mainly due to its randomized procedure. The application of this model revealed a good fit of the forecasted rainfall with the real ones for most of the months in 2006 and 2007. Thus it is a good way to make use of weather information. Finally, empirical relationship between rainfall and machinery field workable probability was provided, and thereby machinery field workable probability can be forecasted with monthly rainfall data. To develop forecasting models of operation cost for farm machinery, it was necessary to forecast different cost items, such as depreciation cost, repair and maintenance cost, fuel consumption cost, labor cost and management cost. Data related to 7 sets of working units of JDT654 tractor was further collected, from which forecasting models for five field operation cost items of farm machinery were derived. Results show that the proposed methods for the development of forecasting models were feasible, with 0.8367 for adjusted sum square and 0.8840 for depreciation cost model relating to repair and maintenance, respectively. Forecasting errors were-2.11% and-5.92% for ploughing implement and rotary tiller, respectively. The average forecasting error for the two implements was-3.88%. Case study show an average operation cost for ploughing implement and roatry tiller powered by tractor 110-90 and tractor JDT-654 were 316.47,139.65 yuan per hectare and 242.24,122.64 yuan per hectare respectively, in the first five years. The average cost of them were 126.62% and 97.52% higher, respectively, while the average operation cost of combine JM-1605 was 168.03 yuan per hectare in the first five years, a value being the least one among machineries with age of 3 to 4 years.The development of a non-linear optimization model for farm machinery systems was base on the provision of a number of, say five, sub-models. Optimization objective of the model was to achieve the least machinery field operation cost, with constraints of the quantity of field work, working hours of working units or combines, field operation sequences, target for select some specific machinery, available working hours and non-negative variables. Thus the main advantages of the model were to allow the use of sub-models to forecast individual cost items, to consider the sequences of machinery field operation, and to select special model machinery etc.. The optimization results were achieved by using LINDO software.In order to evaluate typical farm machinery systems of rice production in the three main rice planting regions in China, i.e. the one-season rice production region of South China(OSSC), two-season rice production region of South China(TSSC) and rice production region of North China(NC), a brief introduction of these typical systems was followed with a series field tests. Rice planting methods include traditional manual transplanter(TMT), rice mechanized transplanter(MT), rice mechanized direct seeding(MDS) and rice mechanized potted-seedling transplanter(MPST). Variables used in the test include 6 types of rice seeds, seed cost, fertilizer and chemical cost, water cost, machinery field operation cost, labour cost, as well as main rice yield indexes of sample fringes, sample grains, sample fruit grains, average weight per thousand grains, average theoretical yields, average real yields, labour hours for each field operation etc.. Based on the data, evaluation were assigned to the effects of labour saving, cost saving and benefit increasing. So that the ranking list for each total operation cost was provided, from high to low, as TMT, MDS and MT. Their average machinery field operation costs account to 28.56%,43.20% and 42.91% in OSSC, respectively. The highest total operational costs were TMT and MT, either for early and for late rice production, with a respective difference of 11.46% and 9.47%. Average machinery field operation costs account to 33.92%,30.72% and 48.88%,43.30% respectively in TSSC. Total operation costs for the system with TMT and MPST are close, their respective average machinery field operation costs account to 18.69% and 25.70% in NC. Compared with the system of TMT, MT was recommended for both OSSC and TSSC. ough rice yields and net profit are increased by 7.53% and 46.78%in average, and labor hours is reduced by 41.44% in average. Rough rice yields are decreased by 6.34% and labor hours is reduced by 31.30% in average for the system with MDS, their net profit are close or increased by 8.5% for traditional rice and cross one respectively. The system with MPST is recommended in NC, compared to the system with TMT. Rough rice yields and net profit were increased by 8.95% and 22.57% respectively, and labor hours reduced by 64.29%. These results served as a good guideline for the selection of farm machinery system for different rice planting regions in China.Finally, some suggestions are also given for further research. |
PDF Full Text Request