中文摘要：

文章利用2001年1月至2012年12月共2912个我国豆粕期货价格时间序列为样本，基于C-C与G-P相结合的方法重构了最佳嵌入维为9且最佳延迟时间为29的相空间，在此基础上得到非整数形式的分形维，并通过小数量法得到正的最大Lyapunov指数，论证了我国豆粕期货市场的混沌特征。研究表明，我国豆粕期货市场并非有效市场，其价格波动兼具随机与确定性，不具备长期预测能力，但可进行周期为251 d的短期预测，这对期货市场价格的预测研究具有很好的借鉴意义。

英文摘要：

According to the sample of soybean meal futures price series in China from Jan 2001to Dec 2012, 2912 data in total, we reconstruct a phase space by method of C-C and G-P. This phase space is with the best embedding dimension for 9 and the optimal delay time for 29. On this basis, a fractal dimension is calculated in non-integer form and the positive maximal Lyapunov exponent is obtained with small-data algorithm method. All prove the chaotic characteristic of the soybean meal futures market in China. Study shows our soybean meal futures market is not a valid one. Its price volatility is both random and certain. Therefore, there is no long time predictive ability available. However, a short-term prediction for 251 days is possible, which can be a good reference of futures market value prediction.