2019数字中国创新大赛混凝土泵车砼活塞故障预警完整源码+说明（亚军）.zip

# DCIC-2019-HNT-2th-Place 2019数字中国创新大赛 混凝土泵车砼活塞故障预警 亚军 #### hnt_concat_data.py： 把所有文件连接成一个文件 #### hnt_features.py： 提取特征 ```python for col in ['发动机转速', '油泵转速', '泵送压力', '液压油温', '流量档位', '分配压力', '排量电流']: with timer(f'make feature for {col}'): fea = df.groupby('sample_file_name')[col].agg([ (f'min_{col}','min'), (f'max_{col}','max'), # (f'median_{col}','median'), (f'mean_{col}','mean'), (f'nuni_{col}','nunique'), # (f'std_{col}','std'), # (f'skew_{col}','skew'), # (f'kurtosis_{col}', agg_kurtosis), # (f'normaltest_{col}',agg_normaltest_pvalue), # (f'num_bins_{col}', agg_num_bins), ]).reset_index() feat_df = feat_df.merge(fea, on=['sample_file_name']) ``` #### lgb.ipynb：5折lgihtgbm #### lgb_fakeid.ipynb：5折lightgbm, 多了个fake_id特征 ```python tmp = feat_df['活塞工作时长'].astype(str)+'#'\ +feat_df['设备类型'].astype(str)+'#'\ +feat_df['低压开关'].astype(str)+'#'\ +feat_df['正泵'].astype(str) lbl = LabelEncoder() feat_df['fake_id'] = lbl.fit_transform(tmp) ``` #### rf.ipynb：5折random forest #### ensemble.ipynb：对不同模型输出的预测概率做加权融合 ```python prob = 0.45*lgb_prob1+0.15*lgb_prob2 + 0.4*rf_prob sub = test_feat_df[['sample_file_name']].copy() sub['label'] = (prob>0.4588).astype(int) sub.label.value_counts() # 输出 # 0 28321 # 1 23929 # Name: label, dtype: int64 ``` 选择这个奇怪的阈值0.4588的原因是，在A榜时发现23900左右个1分数最佳，所以后面一直取一个概率阈值，使得1的个数在23900左右。这个概率通常在0.46左右