From 3180878150f6e22b30394f745008d81cb80af12d Mon Sep 17 00:00:00 2001
From: Jay <csj123456>
Date: 星期四, 17 四月 2025 13:58:02 +0800
Subject: [PATCH] 新增累计真实值和累计预测值的导出
---
iailab-module-model/iailab-module-model-biz/src/main/java/com/iailab/module/model/mcs/pre/service/impl/MmPredictInfluenceFactorServiceImpl.java | 302 ++++++++++++++++++++++++++++++++++++++++++++++++++
1 files changed, 302 insertions(+), 0 deletions(-)
diff --git a/iailab-module-model/iailab-module-model-biz/src/main/java/com/iailab/module/model/mcs/pre/service/impl/MmPredictInfluenceFactorServiceImpl.java b/iailab-module-model/iailab-module-model-biz/src/main/java/com/iailab/module/model/mcs/pre/service/impl/MmPredictInfluenceFactorServiceImpl.java
new file mode 100644
index 0000000..cc97a76
--- /dev/null
+++ b/iailab-module-model/iailab-module-model-biz/src/main/java/com/iailab/module/model/mcs/pre/service/impl/MmPredictInfluenceFactorServiceImpl.java
@@ -0,0 +1,302 @@
+package com.iailab.module.model.mcs.pre.service.impl;
+
+import com.baomidou.mybatisplus.core.metadata.IPage;
+import com.iailab.framework.common.pojo.PageResult;
+import com.iailab.framework.common.service.impl.BaseServiceImpl;
+import com.iailab.framework.common.util.date.DateUtils;
+import com.iailab.module.data.api.point.DataPointApi;
+import com.iailab.module.data.api.point.dto.ApiPointDTO;
+import com.iailab.module.data.api.point.dto.ApiPointValueDTO;
+import com.iailab.module.data.api.point.dto.ApiPointValueQueryDTO;
+import com.iailab.module.data.enums.DataPointFreqEnum;
+import com.iailab.module.model.api.mcs.dto.MmPredictInfluenceFactorHandleReqVO;
+import com.iailab.module.model.mcs.pre.dao.MmPredictInfluenceFactorDao;
+import com.iailab.module.model.mcs.pre.entity.MmPredictInfluenceFactorConfigEntity;
+import com.iailab.module.model.mcs.pre.entity.MmPredictInfluenceFactorResultEntity;
+import com.iailab.module.model.mcs.pre.enums.InfluenceFactorPatternEnum;
+import com.iailab.module.model.mcs.pre.service.MmItemResultJsonService;
+import com.iailab.module.model.mcs.pre.service.MmPredictInfluenceFactorResultService;
+import com.iailab.module.model.mcs.pre.service.MmPredictInfluenceFactorService;
+import com.iailab.module.model.mcs.pre.vo.*;
+import lombok.extern.slf4j.Slf4j;
+import org.apache.commons.lang3.StringUtils;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.stereotype.Service;
+import org.springframework.util.CollectionUtils;
+
+import java.math.BigDecimal;
+import java.util.*;
+import java.util.function.BinaryOperator;
+import java.util.function.Function;
+import java.util.stream.Collectors;
+
+@Slf4j
+@Service
+public class MmPredictInfluenceFactorServiceImpl extends BaseServiceImpl<MmPredictInfluenceFactorDao, MmPredictInfluenceFactorConfigEntity> implements MmPredictInfluenceFactorService {
+
+ @Autowired
+ private MmItemResultJsonService mmItemResultJsonService;
+
+ @Autowired
+ private DataPointApi dataPointApi;
+
+ @Autowired
+ private MmPredictInfluenceFactorResultService influenceFactorResultService;
+
+ @Override
+ public PageResult<MmPredictInfluenceFactorConfigVO> page(MmPredictInfluenceFactorPageReqVO params) {
+ IPage<MmPredictInfluenceFactorConfigVO> page = baseDao.selectPage(params);
+ return new PageResult<>(page.getRecords(), page.getTotal());
+ }
+
+ @Override
+ public MmPredictInfluenceFactorConfigVO getInfo(String id) {
+ return baseDao.getInfo(id);
+ }
+
+ @Override
+ public void create(MmPredictInfluenceFactorConfigVO vo) {
+ MmPredictInfluenceFactorConfigEntity entity = new MmPredictInfluenceFactorConfigEntity();
+ String configId = UUID.randomUUID().toString();
+ entity.setId(configId);
+ entity.setOutputId(vo.getOutputId());
+ entity.setPattern(vo.getPattern());
+ entity.setIsEnable(vo.getIsEnable());
+ entity.setCreateTime(new Date());
+ baseDao.insert(entity);
+
+ // 添加因素
+ if (CollectionUtils.isEmpty(vo.getInfluenceFactors())) {
+ return;
+ }
+ List<MmPredictInfluenceFactorVO> influenceFactors = vo.getInfluenceFactors();
+ influenceFactors.forEach(e -> {
+ e.setId(UUID.randomUUID().toString());
+ e.setConfigId(configId);
+ });
+
+ baseDao.insertInfluenceFactor(influenceFactors);
+ }
+
+ @Override
+ public void update(MmPredictInfluenceFactorConfigVO vo) {
+ MmPredictInfluenceFactorConfigEntity entity = new MmPredictInfluenceFactorConfigEntity();
+ entity.setId(vo.getId());
+ entity.setOutputId(vo.getOutputId());
+ entity.setPattern(vo.getPattern());
+ entity.setIsEnable(vo.getIsEnable());
+ baseDao.updateById(entity);
+
+ // 删除因素
+ baseDao.deleteInfluenceFactor(vo.getId());
+ // 添加因素
+ if (CollectionUtils.isEmpty(vo.getInfluenceFactors())) {
+ return;
+ }
+ List<MmPredictInfluenceFactorVO> influenceFactors = vo.getInfluenceFactors();
+ influenceFactors.forEach(e -> {
+ if (StringUtils.isBlank(e.getId())) {
+ e.setId(UUID.randomUUID().toString());
+ }
+ e.setConfigId(vo.getId());
+ });
+
+ baseDao.insertInfluenceFactor(influenceFactors);
+
+ }
+
+ @Override
+ public void delete(String id) {
+ // 删除因素
+ baseDao.deleteById(id);
+ }
+
+ @Override
+ public void influenceFactorHandle(MmPredictInfluenceFactorHandleReqVO reqVO) {
+ // 影响时间
+ Date influenceTime = new Date(reqVO.getInfluenceTime());
+ // 判断影响时间一定是历史时间
+ if (influenceTime.after(new Date())) {
+ log.info("计算影响因素失败：影响时间大于当前时间");
+ return;
+ }
+ // 查询所有影响因素
+ Map<String, Object> params = new HashMap<>();
+ params.put("isEnable",1);
+ List<MmPredictInfluenceFactorHandleVO> influenceFactorConfigs = baseDao.selectList(params);
+
+ // 根据统计方式分组
+ Map<String, List<MmPredictInfluenceFactorHandleVO>> patternMap = influenceFactorConfigs.stream().collect(Collectors.groupingBy(MmPredictInfluenceFactorHandleVO::getPattern));
+ List<MmPredictInfluenceFactorResultEntity> resultList = new ArrayList<>(influenceFactorConfigs.size());
+ for (Map.Entry<String, List<MmPredictInfluenceFactorHandleVO>> entry : patternMap.entrySet()) {
+ List<MmPredictInfluenceFactorHandleVO> influenceFactors = entry.getValue();
+ // 缓存 因素预测项id-结果，相同因素预测项id只用计算一次
+ HashMap<String,Double> values = new HashMap<>();
+ switch (InfluenceFactorPatternEnum.fromCode(entry.getKey())) {
+ case ACCUMULATE:
+ for (MmPredictInfluenceFactorHandleVO influenceFactor : influenceFactors) {
+ Double value = influenceFactorAccumulateCalculate(influenceFactor, values, influenceTime);
+ if (value == null) {
+ continue;
+ }
+ MmPredictInfluenceFactorResultEntity resultEntity = new MmPredictInfluenceFactorResultEntity();
+ resultEntity.setFactorId(influenceFactor.getId());
+ resultEntity.setTime(influenceTime);
+ resultEntity.setValue(value);
+ resultList.add(resultEntity);
+ }
+ break;
+ case ACCURACY:
+ for (MmPredictInfluenceFactorHandleVO influenceFactor : influenceFactors) {
+ Double value = influenceFactorAccuracyCalculate(influenceFactor, values, influenceTime);
+ if (value == null) {
+ continue;
+ }
+ MmPredictInfluenceFactorResultEntity resultEntity = new MmPredictInfluenceFactorResultEntity();
+ resultEntity.setFactorId(influenceFactor.getId());
+ resultEntity.setTime(influenceTime);
+ resultEntity.setValue(value);
+ resultList.add(resultEntity);
+ }
+ break;
+ default:
+ }
+ }
+ if (CollectionUtils.isEmpty(resultList)) {
+ return;
+ }
+ influenceFactorResultService.insert(resultList);
+
+ }
+
+ private Double influenceFactorAccuracyCalculate(MmPredictInfluenceFactorHandleVO influenceFactor, HashMap<String, Double> values, Date influenceTime) {
+ if (values.containsKey(influenceFactor.getFactorOutputId())) {
+ return values.get(influenceFactor.getFactorOutputId());
+ }
+ // 根据 预测长度 和 预测粒度 推算预测时间
+ Calendar calendar = Calendar.getInstance();
+ calendar.setTime(influenceTime);
+ calendar.add(Calendar.SECOND,-1 * influenceFactor.getPredictlength() * influenceFactor.getGranularity());
+ Date predictTime = calendar.getTime();
+ // 获取result_json,预测历史数据
+ List<Object[]> predictData = mmItemResultJsonService.getData(influenceFactor.getFactorOutputId(), predictTime, DateUtils.FORMAT_YEAR_MONTH_DAY_HOUR_MINUTE_SECOND);
+ if (CollectionUtils.isEmpty(predictData)) {
+ return null;
+ }
+ // 获取真实测点历史值
+ ApiPointValueQueryDTO queryDTO = new ApiPointValueQueryDTO();
+ ApiPointDTO point = dataPointApi.getInfoById(influenceFactor.getPointid());
+ queryDTO.setPointNo(point.getPointNo());
+ queryDTO.setStart(predictTime);
+ queryDTO.setEnd(influenceTime);
+ List<ApiPointValueDTO> dataList = dataPointApi.queryPointHistoryValue(queryDTO);
+ if (CollectionUtils.isEmpty(dataList)) {
+ return null;
+ }
+ dataList = completionData(influenceFactor.getPredictlength(),dataList,predictTime,influenceTime,point.getMinfreqid());
+ Map<String, Double> dataMap = dataList.stream().collect(Collectors.toMap(e -> DateUtils.format(e.getT(), DateUtils.FORMAT_YEAR_MONTH_DAY_HOUR_MINUTE_SECOND), ApiPointValueDTO::getV, (e1, e2) -> e1));
+ // 计算精确度误差
+ int sum = 0;
+ for (int i = 0; i < predictData.size(); i++) {
+ Object[] objects = predictData.get(i);
+ if (dataMap.containsKey(objects[0].toString())) {
+ double abs = Math.abs(new BigDecimal(objects[1].toString()).subtract(BigDecimal.valueOf(dataMap.get(objects[0].toString()))).doubleValue());
+ if (abs > influenceFactor.getDeviationValue()) {
+ sum++;
+ }
+ }
+ }
+
+ double value = BigDecimal.valueOf(sum).divide(BigDecimal.valueOf(predictData.size()), 2, BigDecimal.ROUND_HALF_UP).doubleValue();
+
+ values.put(influenceFactor.getFactorOutputId(),value);
+ return value;
+ }
+
+ private Double influenceFactorAccumulateCalculate(MmPredictInfluenceFactorHandleVO influenceFactor, HashMap<String, Double> values,Date influenceTime) {
+ if (values.containsKey(influenceFactor.getFactorOutputId())) {
+ return values.get(influenceFactor.getFactorOutputId());
+ }
+ // 根据 预测长度 和 预测粒度 推算预测时间
+ Calendar calendar = Calendar.getInstance();
+ calendar.setTime(influenceTime);
+ calendar.add(Calendar.SECOND,-1 * influenceFactor.getPredictlength() * influenceFactor.getGranularity());
+ Date predictTime = calendar.getTime();
+ // 获取result_json,预测历史数据
+ double[] predictData = mmItemResultJsonService.getSimpleData(influenceFactor.getFactorOutputId(), predictTime, influenceFactor.getPredictlength());
+ if (predictData == null || predictData.length == 0) {
+ return null;
+ }
+ // 获取真实测点历史值
+ ApiPointValueQueryDTO queryDTO = new ApiPointValueQueryDTO();
+ ApiPointDTO point = dataPointApi.getInfoById(influenceFactor.getPointid());
+ queryDTO.setPointNo(point.getPointNo());
+ queryDTO.setStart(predictTime);
+ queryDTO.setEnd(influenceTime);
+ List<ApiPointValueDTO> dataList = dataPointApi.queryPointHistoryValue(queryDTO);
+ if (CollectionUtils.isEmpty(dataList)) {
+ return null;
+ }
+ dataList = completionData(influenceFactor.getPredictlength(),dataList,predictTime,influenceTime,point.getMinfreqid());
+ // 计算累计量误差
+ double value = Math.abs(dataList.stream().mapToDouble(ApiPointValueDTO::getV).sum() - Arrays.stream(predictData).sum());
+
+ values.put(influenceFactor.getFactorOutputId(),value);
+ return value;
+ }
+
+ private List<ApiPointValueDTO> completionData(int length, List<ApiPointValueDTO> dataList, Date startTime, Date endTime, String minfreqid) {
+ if (CollectionUtils.isEmpty(dataList) || length == dataList.size()) {
+ return dataList;
+ } else if (length < dataList.size()) {
+ return dataList.subList(dataList.size() - length, dataList.size());
+ }
+
+ List<ApiPointValueDTO> result = new ArrayList<>();
+ long start = startTime.getTime();
+ long end = endTime.getTime();
+ long oneMin = 1000L * DataPointFreqEnum.getEumByCode(minfreqid).getValue();
+ long mins = (end - start) / oneMin;
+
+ //找出缺少项
+ Map<Long, Double> sourceDataMap = new HashMap<>(dataList.size());
+ for (ApiPointValueDTO pv : dataList) {
+ sourceDataMap.put(pv.getT().getTime(), pv.getV());
+ }
+
+ Map<Long, Double> dataMap = new LinkedHashMap<>();
+ for (int i = 0; i < mins; i++) {
+ Long key = start + oneMin * i;
+ Double value = sourceDataMap.get(key);
+ dataMap.put(key, value);
+ }
+
+ //补充缺少项
+ int k = 0;
+ Map.Entry<Long, Double> lastItem = null;
+ for (Map.Entry<Long, Double> item : dataMap.entrySet()) {
+ if (k == 0 && item.getValue() == null) {
+ item.setValue(getFirstValue(dataMap));
+ } else if (item.getValue() == null) {
+ item.setValue(lastItem.getValue());
+ }
+ k++;
+ lastItem = item;
+
+ ApiPointValueDTO dataEntity = new ApiPointValueDTO();
+ dataEntity.setT(new Date(item.getKey()));
+ dataEntity.setV(item.getValue());
+ result.add(dataEntity);
+ }
+ return result;
+ }
+
+ private Double getFirstValue(Map<Long, Double> dataMap) {
+ for (Map.Entry<Long, Double> item : dataMap.entrySet()) {
+ if (item.getValue() != null) {
+ return item.getValue();
+ }
+ }
+ return 0.0;
+ }
+}
--
Gitblit v1.9.3