| 摘要 |
Data relating to energy management may be collected and stored from one or more sites. This data may be analyzed by a series of rules, and each rule may look for certain patterns in the data over time. Each time a pattern is detected in the data, the rule may create and store an event back into the database. Each event may represent the detection of a specific condition that starts at a specific time and continues for a specific duration. Each rule may also dynamically assign and update a score to each event that indicates its level of importance and persistence. Once the raw data has been analyzed and events have been created that represent basic conditions, a new set of rules may analyze the events themselves, in addition to, or rather than, the raw data. Successive passes of rules can thus detect higher level, broader problems. |
