"""Project completed SCADA monitoring into an auditable scheduled execution.""" from collections.abc import Sequence from datetime import datetime, timezone import re from uuid import NAMESPACE_URL, uuid5 from app.domain.schemas.periodic_task import PeriodicTaskExecution from app.services.scada_normal_profile import pollution_signal STEP_TITLES = { "scada-diagnosis": ["读取 SCADA 遥测", "检查新鲜度完整率", "诊断压力流量与水质", "判断潜在事件", "整理诊断结论"], "network-simulation": ["冻结模型与边界", "运行水力模型", "比对实测偏差", "定位偏差区域", "输出模型复核"], } def execution_record(document: dict, monitoring: dict, task_key: str) -> dict: return PeriodicTaskExecution( execution_id=uuid5(NAMESPACE_URL, f"{document['replay_id']}:{task_key}:{monitoring['period_end']}"), replay_id=document["replay_id"], task_key=task_key, title="SCADA 数据诊断" if task_key == "scada-diagnosis" else "管网定时模拟", period_start=monitoring["period_start"], period_end=monitoring["period_end"], knowledge_cutoff=monitoring["knowledge_cutoff"], batch_sequence=monitoring["batch_sequence"], status="pending", conclusion="insufficient_data", summary="等待本周期执行。", steps=[{"id": f"step-{index}", "title": title, "status": "pending"} for index, title in enumerate(STEP_TITLES[task_key])], updated_at=datetime.now(timezone.utc), ).model_dump(mode="json") def diagnosis_execution(document: dict, monitoring: dict, pollution_events: Sequence[dict] = ()) -> dict: record = execution_record(document, monitoring, "scada-diagnosis") end = datetime.fromisoformat(monitoring["period_end"]) devices = {device["device_id"]: device for device in document["devices"]} rules = {rule["rule_id"]: rule for rule in document["config"]["rules"]} signals = [] water_events = {} active_events = [state["event"] for state in document.get("rule_states", {}).values() if state.get("event")] for event in [*pollution_events, *active_events]: if not pollution_signal(document, event): continue if event.get("occurred_at") and datetime.fromisoformat(event["occurred_at"]) > end: continue if event.get("batch_sequence", 0) > monitoring["batch_sequence"]: continue water_events[event["event_id"]] = event evidence = {event["event_id"]: event for event in monitoring.get("anomaly_signals", [])} evidence.update(water_events) for signal in evidence.values(): rule = rules.get(signal["rule_id"], {}) water = pollution_signal(document, signal) signals.append({"kind": "suspected_pollution" if water else "hydraulic_anomaly", "device_ids": signal["device_ids"], "detail": signal["detail"], "metric_name": rule.get("metric_name", ""), "rule_id": signal["rule_id"], **({"task_id": signal["task_id"]} if water and signal.get("task_id") else {})}) if monitoring["status"] == "suspected_burst": signals.append({"kind": "suspected_burst", "detail": "压力与爆管侦测等联合证据达到项目嫌疑门槛。", "device_ids": monitoring.get("pressure", {}).get("hit_device_ids", [])}) missing, invalid, fresh_ids = [], [], set() for device_id, device in devices.items(): row = document.get("latest", {}).get(device_id) frequency = str(device.get("transmission_frequency") or "") clock = re.fullmatch(r"(\d+):(\d+):(\d+)", frequency) minutes = re.fullmatch(r"(\d+)\s*(?:min|分钟)", frequency) seconds = (int(clock[1]) * 3600 + int(clock[2]) * 60 + int(clock[3]) if clock else int(minutes[1]) * 60 if minutes else document["monitoring_interval_seconds"]) age = (end - datetime.fromisoformat(row["time"])).total_seconds() if row else None if age is None or age < 0 or age >= 2 * max(seconds, 1): missing.append(device_id) elif row["quality"] not in {"valid", "corrected"}: invalid.append(device_id) else: fresh_ids.add(device_id) conclusion = ("suspected_burst" if monitoring["status"] == "suspected_burst" else "suspected_pollution" if any(signal["kind"] == "suspected_pollution" for signal in signals) else "attention" if signals or monitoring["status"] == "attention" or invalid else monitoring["status"]) gaps = [] if missing: gaps.append("缺少及时观测的测点:" + "、".join(missing)) if invalid: gaps.append("无效读数测点:" + "、".join(invalid)) if monitoring["status"] == "insufficient_data": gaps.append(monitoring["detail"]) water_ids = {key for key, device in devices.items() if device.get("device_type") in {"quality", "water_quality"}} configured_water = {device for rule in rules.values() if rule.get("kind") == "bounds" and rule.get("metric_name") for device in rule.get("device_ids", [])} if water_ids - configured_water: gaps.append("尚未配置明确水质指标与业务阈值的测点,污染风险未评估:" + "、".join(sorted(water_ids - configured_water))) if gaps and conclusion == "normal": conclusion = "insufficient_data" summary = ("发现水质异常,存在污染风险,已进入原因核验。" if conclusion == "suspected_pollution" else monitoring["detail"]) coverage = len(fresh_ids) / len(devices) * 100 if devices else 0 metrics = [{"id": "usable-coverage", "label": "及时有效测点占比", "value": f"{coverage:.1f}", "unit": "%", "status": "attention" if missing or invalid else "normal", "detail": "按设备采样周期检查;缺测或无效读数单列。"}, {"id": "risk-signals", "label": "运行异常信号", "value": str(len(signals)), "unit": "项", "status": "attention" if signals else "normal", "detail": "压力、流量与水质分别判别,异常不等于现场确诊。"}] for key, title, kinds in (("pressure", "压力", {"pressure"}), ("flow", "流量", {"flow", "pipe_flow"}), ("water", "水质", {"quality", "water_quality"})): ids = {device for device, metadata in devices.items() if metadata.get("device_type") in kinds} if ids: valid_count = len(ids & fresh_ids) metrics.append({"id": key + "-freshness", "label": title + "测点及时有效率", "value": f"{valid_count / len(ids) * 100:.1f}", "unit": "%", "status": "normal" if valid_count == len(ids) else "attention", "detail": f"有效 {valid_count} / {len(ids)} 个测点;按设备采样周期核验。"}) weighted = monitoring.get("weighted") if weighted: metrics.append({"id": "burst-score", "label": "爆管嫌疑综合评分", "value": f"{weighted['score']:.2f}", "unit": "", "status": "attention" if weighted["score"] >= weighted["threshold"] else "normal", "detail": f"项目嫌疑门槛 {weighted['threshold']:.2f};综合分不代表概率。"}) record.update(status="completed", conclusion=conclusion, summary=summary, metrics=metrics, data_gaps=gaps, risk_signals=signals, investigation_task_id=monitoring.get("task_id") or next( (signal["task_id"] for signal in signals if signal.get("task_id")), None), evidence=[f"周期 {monitoring['period_start']} 至 {monitoring['period_end']}", f"冻结批次 {monitoring['batch_sequence']};监测回执 {monitoring['monitoring_id']}", *[signal["detail"] for signal in signals]], result={"monitoring": monitoring, "missing_device_ids": missing, "invalid_device_ids": invalid}) for step in record["steps"]: step.update(status="completed") return PeriodicTaskExecution.model_validate(record).model_dump(mode="json")