Start-stop data intelligent analysis system and method for water-turbine generator set

Abstract

The invention discloses an intelligent analysis system and method for startup and shutdown data of a water-turbine generator set, and relates to the technical field of startup and shutdown data analysis of the water-turbine generator set, and the system comprises an operation data collection module which is used for collecting key operation data in the startup and shutdown process of the set, and forming analysis report data of startup and shutdown of the set at each time; the first early warning module is used for analyzing the data change in the working condition conversion when the unit is started and stopped, and generating first early warning information when the data change amplitude does not meet a preset threshold value; the second early warning module is used for acquiring a distribution interval of the unit equipment operation data and generating second early warning information when the unit equipment operation data deviates from the distribution interval; and the association analysis module is used for judging an association relationship between the first early warning information and the second early warning information based on the first early warning information and the second early warning information, and storing the association relationship in a database. According to the method, the equipment influencing the startup and shutdown process can be analyzed, the startup and shutdown time is shortened, and the startup and shutdown success rate is increased.

Description

Technical Field

[0001] The invention relates to the technical field of start-up and shutdown data analysis of a hydro-generator set, and in particular to a start-up and shutdown data intelligent analysis system and method for a hydro-generator set. Background Art

[0002] In the current new power system, the proportion of wind and solar power generation is increasing. Large hydropower units are gradually transformed into peak-shaving and frequency-regulating power stations due to their unique characteristics of fast response and regulation, playing an important role in the stability of the power grid. However, due to frequent and deeper peak-shaving and frequency-regulating tasks, frequent start-up and shutdown of units have become the norm, and the operation and maintenance management of equipment are facing severe challenges.

[0003] At present, the start-up and shutdown data analysis of hydro-turbine generator sets mainly relies on manual analysis based on expert experience, which has high requirements on personnel and cannot be analyzed in real time and dynamically. With the rise of advanced technologies such as artificial intelligence and machine learning, how to rely on the industrial Internet platform, introduce advanced algorithms such as artificial intelligence and machine learning, and make intelligent judgments and analyses on auxiliary equipment that affect the start-up and shutdown process, reduce the start-up and shutdown time, and improve the success rate of start-up and shutdown has become one of the problems that need to be solved urgently. Summary of the invention

[0004] The purpose of the present invention is to provide a system and method for intelligent analysis of start-up and shutdown data of a hydro-generator set, so as to solve the problems raised in the prior art.

[0005] To achieve the above object, the present invention provides the following technical solution: a method for intelligently analyzing the start-up and shutdown data of a hydro-generator set, the method comprising:

[0006] S1. Collect key operating data during the start-up and shutdown of the unit, and form analysis report data for each start-up and shutdown of the unit;

[0007] S2, analyzing the data changes during the operation state conversion of the unit when it is turned on and off, and generating the first warning information when the data change amplitude does not meet the predetermined threshold;

[0008] S3, obtaining the distribution interval of the unit equipment operation data, and generating a second warning information when the unit equipment operation data deviates from the distribution interval;

[0009] S4. Based on the first warning information and the second warning information, determine the correlation between the first warning information and the second warning information, and store the correlation in a database.

[0010] According to the above technical solution, the key operating data include: unit start-up and shutdown time consumption data, unit start-up and shutdown parameter data, unit start-up and shutdown bearing bearing temperature data and unit start-up and shutdown vibration data.

[0011] According to the above technical solution, the unit start-up and shutdown time-consuming data refers to the time data used for a single unit start-up and shutdown under the same working condition conversion; specifically including: the time from when the unit starts to rotate to when the unit speed reaches the rated value during the start-up process, the time when the guide vane opening change reaches the steady state value, and the time from when the guide vane starts to act to the first entry into the synchronization band; the time when the unit speed drops to the set speed during the shutdown process, and the time when the guide vane drops from the opening to zero; among them, the rated value, steady state, synchronization band, and set speed all adopt the set values of the system staff;

[0012] The unit start-up and shutdown parameter data refers to the change data of the parameters during a single unit start-up and shutdown process, specifically including: the swing data of the guide vane and the pause data during the swing of the guide vane;

[0013] The unit start-up and shutdown bearing pad temperature data refers to the bearing pad temperature, oil temperature, and cooling water temperature data of each part of the unit during the start-up and shutdown process;

[0014] The unit start-up and shutdown vibration data refers to the vibration, swing, and pressure pulsation signal data of the unit during the start-up and shutdown process.

[0015] According to the above technical solution, the working condition conversion specifically includes: conversion from shutdown to idling working condition, conversion from shutdown to no-load working condition, conversion from shutdown to generating working condition, conversion from idling to no-load working condition, conversion conditions from idling to generating working condition, conversion from no-load to generating working condition, conversion from generating to no-load working condition, conversion from generating to idling working condition, conversion from generating to shutdown working condition, conversion from no-load to idling working condition, conversion from no-load to shutdown working condition, and conversion from idling to shutdown working condition.

[0016] According to the above technical solution, analyzing the data change of the unit equipment during the working condition conversion includes:

[0017] Taking N groups of unit start-up and shutdown time-consuming data under any working condition conversion as the historical reference data group under this working condition conversion, among which, the calculation of the predetermined threshold includes:

[0018]

[0019] Among them, h min and h max respectively represent the upper and lower limit ranges of the predetermined threshold; h0 represents the average value of N groups of unit start-up and shutdown time-consuming data; h1 and h2 respectively represent the minimum and maximum values of N groups of unit start-up and shutdown time-consuming data; N refers to a constant term;

[0020] If there is unit start-up and shutdown time-consuming data under this working condition conversion that is lower than h min or higher than h max , it is determined that the predetermined threshold is not satisfied.

[0021] According to the above technical solution, judging that the unit equipment operation data deviates from the distribution interval includes:

[0022] Obtain the distribution range of the operating data of the unit equipment, specifically including:

[0023] For any one of the operating data of the unit equipment, such as the unit start-stop parameter data, the unit start-stop bearing pad temperature data, and the unit start-stop vibration data, establish a two-dimensional reference point corresponding to the data, and the format is written as (x0, y0), where x0 represents the acquisition batch of the data; y0 represents the actual number of the data;

[0024] Obtain the corresponding two-dimensional reference point of the operating data of the unit equipment to be judged, and at the same time take the corresponding two-dimensional reference points of the corresponding N groups of historical data to form a judgment set;

[0025] Set the nearest neighbor point parameter k for the judgment set. For each point in the judgment set, calculate its distance to other points, and find the nearest k points, and calculate the local density of the point to be measured:

[0026]

[0027] Among them, L p represents the local density of the point p to be measured; m represents the set formed by the nearest k points of the point p to be measured; q represents any point in the set m; D(p, q) represents the reachable distance between the point p and the point q, and the calculation of the reachable distance includes:

[0028] D(p, q) = max(k distance、q , d(p, q))

[0029] Among them, k distance、q represents the distance from the point q to its k-th nearest neighbor; d(p, q) represents the actual distance between the point p and the point q;

[0030] Calculate the local outlier factor of the point p to be measured:

[0031]

[0032] Among them, L q represents the local density of the point q; LOF(p) represents the local outlier factor of the point p to be measured;

[0033] If the local outlier factor of the point p to be measured is the highest among the local outlier factors of the N groups of historical data, it is determined that the operating data of the unit equipment corresponding to the point p to be measured deviates from the distribution range.

[0034] According to the above technical solution, the association relationship between the first warning information and the second warning information includes:

[0035] Take N groups of warning data. In any one group of warning data, there is at least one type of warning information. Calculate the prior probabilities of the existence of the second warning information, denoted as G1 and G2 respectively. Among them, G1 refers to the prior probability of the existence of the second warning information, and G2 refers to the prior probability of the non-existence of the second warning information; calculate the probability G3 that the first warning information does not exist on the premise that the second warning information exists; calculate the probability G4 that the first warning information does not exist on the premise that the second warning information does not exist;

[0036] Calculate the marginal probability G5 = G1 * G3 + G2 * G4;

[0037] Based on the marginal probability G5, form the probability G6 that the second warning information exists on the premise that the first warning information does not exist:

[0038]

[0039] Based on the probability G6, form a reference point for equipment maintenance. Continuously supplement new data into the N groups of warning data in real time, and continuously solve for the probability G6. If the probability G6 increases, the equipment maintenance interval time decreases; if the probability G6 decreases, the equipment maintenance interval time increases.

[0040] An intelligent analysis system for start-stop data of hydro-generator sets, which includes:

[0041] An operation data acquisition module, which is used to collect key operation data during the start-stop process of the unit and form analysis report data for each start-stop of the unit;

[0042] A first warning module, which is used to analyze the data changes during the working condition conversion of the unit start-stop. When the data change amplitude does not meet the predetermined threshold, a first warning information is generated;

[0043] A second warning module, which is used to obtain the distribution interval of the operation data of the unit equipment. When the operation data of the unit equipment deviates from the distribution interval, a second warning information is generated;

[0044] A correlation analysis module, based on the first warning information and the second warning information, judges the correlation relationship between the first warning information and the second warning information and stores it in the database.

[0045] According to the above technical solution, based on the correlation relationship between the first warning information and the second warning information, form a reference point for equipment maintenance of the unit. Continuously supplement new data into the training process of the warning data group in real time, and continuously solve. If the solved value increases, the equipment maintenance interval time decreases; if the solved value decreases, the equipment maintenance interval time increases.

[0046] Compared with the prior art, the beneficial effects of the present invention are as follows: Based on the collection and analysis of key influencing factors during the startup and shutdown processes of the unit, the present application realizes the extraction of key operating data characteristics during the startup and shutdown processes of the unit. By using advanced technologies such as artificial intelligence and machine learning, the key operating data for each startup and shutdown is analyzed and processed to form data comparisons under different condition conversion conditions. It can effectively judge the fault situation of the unit based on the data changes during the condition conversion of the unit's startup and shutdown and the distribution range of the unit equipment operating data, thereby reducing the startup and shutdown time and increasing the success rate of startup and shutdown. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic flow chart of the intelligent analysis method for startup and shutdown data of the hydro-generator unit of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0049] Embodiment: As Figure 1 shown, the present invention provides an intelligent analysis method for startup and shutdown data of a hydro-generator unit, and the method includes:

[0050] Collect key operating data during the startup and shutdown processes of the unit to form analysis report data for each startup and shutdown of the unit;

[0051] The key operating data includes: startup and shutdown time data of the unit, startup and shutdown parameter data of the unit, bearing pad temperature data during startup and shutdown of the unit, and vibration data during startup and shutdown of the unit.

[0052] The startup and shutdown time data of the unit refers to the time data used for a single startup and shutdown of the unit under the same condition conversion; specifically includes: the time from when the unit starts to rotate to when the unit speed reaches the rated value during the startup process, the time when the guide vane opening change reaches the steady state value, and the time from when the guide vane starts to act to the first entry into the synchronization band; the time when the unit speed drops to the specified speed during the shutdown process, and the time when the guide vane drops from the opening to zero; wherein, the rated value, steady state, synchronization band, and specified speed all adopt the set values of the system staff;

[0053] The unit startup and shutdown parameter data refers to the data of parameter changes during a single unit startup and shutdown process, specifically including: the swing data of the guide vane and the pause data during the swing of the guide vane; the unit startup and shutdown bearing pad temperature data refers to the bearing pad temperature, oil temperature, and cooling water temperature data of each part of the unit during startup and shutdown; the unit startup and shutdown vibration data refers to the vibration, swing, and pressure pulsation signal data of the unit during startup and shutdown. Data cleaning is to screen and remove duplicate and redundant data, supplement missing data, correct or delete incorrect data, and finally organize it into data that we can further process and use.

[0054] In this embodiment, it is developed and applied relying on the industrial Internet platform to intelligently collect and analyze the data and equipment related to the key factors of unit startup and shutdown. The hardware resources are as shown in Table 1 below:

[0055] Table 1: Hardware Resources on the Substation Side

[0056]

[0057] Among them, the virtual node resources are as shown in Table 2 below:

[0058] Table 2: Virtual Node Resources

[0059]

[0060] In the process of data collection, for the handling of missing data values, the following methods are usually used: deleting missing values; when the number of samples is large and the proportion of samples with missing values in the entire sample is relatively small, in this case, we can use the simplest and most effective method to handle the missing value situation. That is to directly discard the samples with missing values.

[0061] Mean filling method: Divide the data into several groups according to the attribute with the largest correlation coefficient with the missing value attribute, then calculate the mean value of each group respectively, and put these mean values into the missing values.

[0062] Hot deck imputation method: Find a most similar object in the database, and then use the value of this similar object for filling.

[0063] The specific conditions of the operating condition conversion include: conversion from shutdown to idling condition, conversion from shutdown to no-load condition, conversion from shutdown to generating condition, conversion from idling to no-load condition, conversion from idling to generating condition, conversion from no-load to generating condition, conversion from generating to no-load condition, conversion from generating to idling condition, conversion from generating to shutdown condition, conversion from no-load to idling condition, conversion from no-load to shutdown condition, and conversion from idling to shutdown condition.

[0064] The specific start-up and shutdown processes are as follows: Start-up process: Shutdown state → Turn on the technical water supply of the unit → Cut off the foundation pit heater → Turn on the generator oil mist absorption device → Turn on the carbon powder vacuum cleaner → Start the hydraulic system → Turn on the high-pressure oil → Pull out the servomotor lock → Start up to the governor → Idle state → Close the field discharge switch → Apply excitation → No-load state → Close the generator outlet circuit breaker → Power generation state.

[0065] The start-up process includes 6 sub-processes: "Shutdown to idle", "Shutdown to no-load", "Shutdown to power generation", "Idle to no-load", "Idle to power generation", and "No-load to power generation".

[0066] When the unit is in one of the states of "shutdown", "idle", or "no-load" and the corresponding operating conditions are met, a start-up command can be issued to the unit.

[0067] The shutdown process is specifically as follows: Power generation state → Reduce the load to zero → Trip the generator outlet circuit breaker → Shutdown to excitation → Shutdown to governor → Turn on the high-pressure oil → Apply electric braking → Check the guide vane position → Check the unit speed → Turn on the brake and dust collector → Check the unit speed → Insert the servomotor lock → Cut off the brake → Close the technical water supply and main shaft seal water of the unit → Turn on the foundation pit heater → Cut off the generator oil mist absorption device → Cut off the carbon powder vacuum cleaner → Cut off the brake dust collector → Turn on the creep monitoring device → Shutdown state.

[0068] Analyze the data changes during the working condition conversion of the unit start-up and shutdown. When the data change amplitude does not meet the predetermined threshold, a first warning message is generated;

[0069] The analysis of the data changes of the unit equipment during the working condition conversion includes:

[0070] Take N groups of unit start-up and shutdown time-consuming data under any working condition conversion as the historical reference data group under this working condition conversion. Among them, the calculation of the predetermined threshold includes:

[0071]

[0072] Among them, h min and h max respectively represent the upper and lower limit ranges of the predetermined threshold; h0 represents the average value of the N groups of unit start-up and shutdown time-consuming data; h1 and h2 respectively represent the minimum and maximum values of the N groups of unit start-up and shutdown time-consuming data; N refers to a constant term;

[0073] If there is unit start-up and shutdown time-consuming data under this working condition conversion that is lower than h min or higher than h max , it is determined that the predetermined threshold is not met.

[0074] Obtain the distribution interval of the unit equipment operation data. When the unit equipment operation data deviates from the distribution interval, a second warning message is generated;

[0075] Determining that the operation data of the unit equipment deviates from the distribution range includes:

[0076] Obtaining the distribution range of the operation data of the unit equipment, specifically including:

[0077] For any one of the operation data of the unit equipment, such as the unit start-stop parameter data, the unit start-stop bearing pad temperature data, and the unit start-stop vibration data, establish a two-dimensional reference point corresponding to the data, and the format is written as (x0, y0), where x0 represents the acquisition batch of the data; y0 represents the actual number of the data.

[0078] Obtain the corresponding two-dimensional reference point of the operation data of the unit equipment to be judged, and at the same time take the corresponding two-dimensional reference points of the corresponding N groups of historical data to form a judgment set.

[0079] Set the nearest neighbor point parameter k for the judgment set. For each point in the judgment set, calculate its distance to other points and find the nearest k points, and calculate the local density of the point to be measured:

[0080]

[0081] Among them, L p represents the local density of the point p to be measured; m represents the set formed by the nearest k points of the point p to be measured; q represents any point in the set m; D(p, q) refers to the reachable distance between the point p and the point q, and the calculation of the reachable distance includes:

[0082] D(p, q) = max(k distance、q , d(p, q))

[0083] Among them, k distance、q represents the distance from the point q to its kth nearest neighbor; d(p, q) represents the actual distance between the point p and the point q.

[0084] Calculate the local outlier factor of the point p to be measured:

[0085]

[0086] Among them, L q represents the local density of the point q; LOF(p) represents the local outlier factor of the point p to be measured.

[0087] If the local outlier factor of the point p to be measured is the highest among the local outlier factors of the N groups of historical data, it is determined that the operation data of the unit equipment corresponding to the point p to be measured deviates from the distribution range.

[0088] Based on the first warning information and the second warning information, judge the correlation relationship between the first warning information and the second warning information and store it in the database.

[0089] The determination of the correlation between the first warning information and the second warning information includes:

[0090] Select N groups of warning data. In any group of warning data, there is at least one type of warning information. Calculate the prior probabilities of the existence of the second warning information respectively, denoted as G1 and G2. Among them, G1 refers to the prior probability of the existence of the second warning information, and G2 refers to the prior probability of the non-existence of the second warning information; calculate the probability G3 of the non-existence of the first warning information on the premise of the existence of the second warning information; calculate the probability G4 of the non-existence of the first warning information on the premise of the non-existence of the second warning information.

[0091] Calculate the marginal probability G5 = G1 * G3 + G2 * G4;

[0092] Based on the marginal probability G5, form the probability G6 of the existence of the second warning information on the premise of the non-existence of the first warning information:

[0093]

[0094] Based on the probability G6, form a reference point for equipment maintenance. Continuously supplement new data into the N groups of warning data in real time, and continuously solve the probability G6. If the probability G6 increases, the equipment maintenance interval time decreases; if the probability G6 decreases, the equipment maintenance interval time increases.

[0095] It also includes an intelligent analysis system for start-stop data of hydro-generator units. This system includes:

[0096] An operation data acquisition module, which is used to collect key operation data during the start-stop process of the unit and form analysis report data for each start-stop of the unit;

[0097] A first warning module, which is used to analyze the data changes during the working condition conversion of the unit start-stop. When the data change amplitude does not meet the predetermined threshold, it generates the first warning information;

[0098] A second warning module, which is used to obtain the distribution interval of the operation data of the unit equipment. When the operation data of the unit equipment deviates from the distribution interval, it generates the second warning information;

[0099] A correlation analysis module, which, based on the first warning information and the second warning information, determines the correlation between the first warning information and the second warning information and stores it in the database.

[0100] Based on the correlation between the first warning information and the second warning information, form a reference point for unit equipment maintenance. Continuously supplement new data into the training process of the warning data group in real time and continuously solve. If the solution value increases, the equipment maintenance interval time decreases; if the solution value decreases, the equipment maintenance interval time increases.

[0101] It is obvious to those skilled in the art that the present invention is not limited to the details of the above-described exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be embraced within the present invention. Any reference signs in the claims should not be construed as limiting the claims involved.

Claims

1. An intelligent analysis method for start-up and shutdown data of a hydro-generating unit, characterized in that: The method includes: S1. Collect key operation data during the startup and shutdown processes of the unit to form analysis report data for each startup and shutdown of the unit; S2. Analyze the data changes during the working condition conversion of the unit startup and shutdown. When the data change range does not meet the predetermined threshold, generate a first warning message; S3. Obtain the distribution range of the operation data of the unit equipment. When the operation data of the unit equipment deviates from the distribution range, generate a second warning message; S4. Based on the first warning message and the second warning message, judge the correlation relationship between the first warning message and the second warning message, and store it in the database.

2. The intelligent analysis method for start-up and shutdown data of a hydro-generating unit according to claim 1, wherein: The key operation data includes: the startup and shutdown time-consuming data of the unit, the startup and shutdown parameter data of the unit, the bearing pad temperature data of the unit during startup and shutdown, and the vibration data of the unit during startup and shutdown.

3. The intelligent analysis method for startup and shutdown data of a hydro-generator set according to claim 2, characterized in that: The startup and shutdown time-consuming data of the unit refers to the time data used for a single startup and shutdown of the unit under the same working condition conversion; specifically includes: the time from when the unit starts to rotate to when the unit speed reaches the rated value during the startup process, the time when the guide vane opening change reaches the steady state value, and the time from when the guide vane starts to act to the first entry into the synchronization band; the time when the unit speed drops to the specified speed during the shutdown process, and the time when the guide vane drops from the opening to zero; among them, the rated value, steady state, synchronization band, and specified speed all adopt the set values of the system staff; The startup and shutdown parameter data of the unit refers to the change data of the parameters during a single startup and shutdown process of the unit, specifically including: the swing data of the guide vane and the pause data during the swing process of the guide vane; The bearing pad temperature data of the unit during startup and shutdown refers to the bearing pad temperature, oil temperature, and cooling water temperature data of each part of the unit during startup and shutdown; The vibration data of the unit during startup and shutdown refers to the vibration, swing, and pressure pulsation signal data of the unit during startup and shutdown.

4. The intelligent analysis method for start-up and shutdown data of a hydro-generating unit according to claim 1, characterized in that: The working condition conversion specifically includes: the conversion from shutdown to idling working condition, the conversion from shutdown to no-load working condition, the conversion from shutdown to generating working condition, the conversion from idling to no-load working condition, the conversion conditions from idling to generating working condition, the conversion from no-load to generating working condition, the conversion from generating to no-load working condition, the conversion from generating to idling working condition, the conversion from generating to shutdown working condition, the conversion from no-load to idling working condition, the conversion from no-load to shutdown working condition, and the conversion from idling to shutdown working condition.

5. The intelligent analysis method for start-up and shutdown data of a hydro-generating unit according to claim 1, characterized in that: The analysis of the data changes of the unit equipment during the working condition conversion includes: Take N groups of startup and shutdown time-consuming data of the unit under any working condition conversion as the historical reference data group under this working condition conversion, where the calculation of the predetermined threshold includes: Among them, h min and h max respectively represent the upper and lower limit ranges of a predetermined threshold; h0 represents the average value of the start-up and shutdown time-consuming data of N sets of units; h1 and h2 respectively represent the minimum and maximum values of the start-up and shutdown time-consuming data of N sets of units; N refers to a constant term; If the unit start-up and shutdown time data under this operating condition conversion is lower than h min or higher than h max , it is determined that the predetermined threshold is not met.

6. The intelligent analysis method for start-up and shutdown data of a hydro-generating unit according to claim 5, characterized in that: Judging that the operation data of the unit equipment deviates from the distribution range includes: Obtaining the distribution range of the operation data of the unit equipment, specifically including: For any one of the operation data of the unit equipment, such as the startup and shutdown parameter data of the unit, the bearing pad temperature data of the unit during startup and shutdown, and the vibration data of the unit during startup and shutdown, establish a corresponding two-dimensional reference point for the data, and the format is written as (x0, y0), where x0 represents the acquisition batch of the data; y0 represents the actual number of the data; Obtain the corresponding two-dimensional reference point of the operation data of the unit equipment to be judged, and at the same time take the corresponding two-dimensional reference points of the corresponding N groups of historical data to form a judgment set; Set the nearest neighbor parameter k for the judgment set. For each point in the judgment set, calculate its distance to other points, find the nearest k points, and calculate the local density of the point to be measured: Among them, L p represents the local density of the point p to be measured; m represents the set formed by the k nearest points of the point p to be measured; q represents any point in the set m; D(p, q) represents the reachable distance between the point p and the point q, and the calculation of the reachable distance includes: D(p, q) = max(k distance、q , d(p, q)) where k distance、q represents the distance from point q to its k-th nearest neighbor; d(p, q) represents the actual distance between point p and point q; Calculate the local outlier factor of the point to be measured p: Among them, L q represents the local density of point q; LOF(p) represents the local outlier factor of the point p to be measured; If there exists a local outlier factor of the point to be measured p that is the highest among the local outlier factors of N groups of historical data, it is determined that the operation data of the unit equipment corresponding to the point to be measured p deviates from the distribution interval.

7. The intelligent analysis method for start-up and shutdown data of a hydro-generating unit according to claim 6, characterized in that: The determination of the correlation relationship between the first warning information and the second warning information includes: Take N groups of warning data. In any group of warning data, there is at least one type of warning information. Calculate the prior probabilities of the existence of the second warning information, denoted as G1 and G2 respectively. Among them, G1 refers to the prior probability of the existence of the second warning information, and G2 refers to the prior probability of the non-existence of the second warning information; calculate the probability G3 of the non-existence of the first warning information on the premise of the existence of the second warning information; calculate the probability G4 of the non-existence of the first warning information on the premise of the non-existence of the second warning information; Calculate the marginal probability G5 = G1 * G3 + G2 * G4; Based on the marginal probability G5, form the probability G6 of the existence of the second warning information on the premise of the non-existence of the first warning information: Based on the probability G6, form a reference point for equipment maintenance. Continuously supplement new data into the training process of the N-group warning data, and continuously solve the probability G6. If the probability G6 increases, the equipment maintenance interval time decreases; if the probability G6 decreases, the equipment maintenance interval time increases.

8. An intelligent analysis system for start-up and shutdown data of a hydro-generating unit, characterized in that: The system includes: An operation data acquisition module, which is used to collect key operation data during the startup and shutdown of the unit and form analysis report data for each startup and shutdown of the unit; A first warning module, which is used to analyze the data change during the working condition conversion of the unit startup and shutdown. When the data change amplitude does not meet the predetermined threshold, it generates the first warning information; A second warning module, which is used to obtain the distribution interval of the unit equipment operation data. When the unit equipment operation data deviates from the distribution interval, it generates the second warning information; A correlation analysis module, which, based on the first warning information and the second warning information, determines the correlation relationship between the first warning information and the second warning information and stores it in the database.

9. The intelligent analysis system for start-up and shutdown data of a hydro-generating unit according to claim 8, wherein: Based on the correlation relationship between the first warning information and the second warning information, form a reference point for equipment maintenance of the unit. Continuously supplement new data into the training process of the warning data group and continuously solve. If the solved value increases, the equipment maintenance interval time decreases; if the solved value decreases, the equipment maintenance interval time increases.

Images