An offshore wind farm operation and maintenance method and system based on digital twinning

Abstract

The application discloses an offshore wind farm operation and maintenance method and system based on digital twinning, and belongs to the technical field of intelligent operation and maintenance of offshore wind farms. Signals required for establishing a digital twinning system are collected at a data sensing level, data fusion is realized through edge computing, a digital twinning model driven by mechanism data is established, the model is activated by signals collected in a physical space, functions such as power prediction, state evaluation, fault early warning, fault diagnosis and life prediction based on digital twinning are realized, and the digital twinning model can be adaptively corrected as the operating state of equipment changes. Through the actual application of the digital twinning system, the number of field personnel can be effectively reduced, the maintenance time can be reduced, the annual average downtime can be reduced, and the unit failure rate can be reduced, so that the offshore wind power can be safely, efficiently and intelligently developed and utilized on a large scale, and the technical level of offshore wind power operation and maintenance is effectively improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent operation and maintenance of offshore wind farms, and particularly relates to an offshore wind farm operation and maintenance method and system based on digital twinning. BACKGROUND

[0002] Offshore wind power has the advantages of rich resources, not occupying land resources, and being close to the load center along the coast, and is the key development direction of future new energy. High-quality development of offshore wind power is of great significance to accelerate the construction of a clean, low-carbon, safe and efficient energy system.

[0003] Due to the harsh environment of tides, typhoons, air currents and lightning faced by offshore wind power, the unit is prone to failure and the maintenance is more difficult than on land. According to statistics, the operation and maintenance cost of offshore wind power accounts for about 40% of the total investment cost of offshore wind power, so it is urgent to carry out digital and intelligent operation and maintenance technology research to improve the technical level of offshore wind power operation and maintenance. SUMMARY

[0004] In order to solve the defects existing in the prior art, the purpose of the present application is to provide an offshore wind farm operation and maintenance method and system based on digital twinning.

[0005] The present application is realized by the following technical solutions:

[0006] An offshore wind farm operation and maintenance method based on digital twinning, comprising:

[0007] S1: collecting operation data of the offshore wind farm;

[0008] S2: using an edge computing method to perform multi-source heterogeneous data fusion processing on the operation data collected in S1;

[0009] S3: constructing a digital twin model of the offshore wind farm using the operation data processed by S2 multi-source heterogeneous data fusion processing;

[0010] S4: transmitting the data collected in the physical space to the digital twin model constructed in S3 to drive each functional model in the digital twin model to perform calculation and analysis;

[0011] S5: feeding back the calculation and analysis results of S4 to the physical space to guide the operation and maintenance of the offshore wind farm.

[0012] Preferably, in S1, the operation data of the offshore wind farm is collected by an integrated monitoring technology.

[0013] Preferably, in S1, the collected operation data includes blade vibration signals, blade load signals, blade video signals, blade sound signals, transmission chain vibration signals, tower vibration signals, tower load signals and SCADA data.

[0014] Preferably, in S2, for the same type of data, fusion is performed at the data layer; and for different types of data, fusion is performed at the feature layer.

[0015] Preferably, in S3, each functional model in the digital twin model of the offshore wind farm is constructed based on a knowledge-data dual-driven method.

[0016] Further preferably, the functional models include a visualization model of the offshore wind farm, a control model of the wind turbine generator, a dynamics model of the wind turbine generator, a fault prediction model of the wind turbine generator, a life assessment model of the key components, an electrical fault diagnosis model of the offshore substation, and a monitoring and early warning model of the submarine cable.

[0017] Preferably, the digital twin model of the offshore wind farm is self-adaptively corrected and iteratively updated through real-time interaction between the digital twin and the physical entity.

[0018] Further preferably, the self-adaptive correction specifically includes arranging vibration sensors in 8 layers in the height direction of the tower to measure the 7th order vibration frequency and the corresponding vibration mode of the tower, and taking the minimum of the actual value and the finite element calculation value of the natural frequency and the vibration mode as the objective function, i.e.

[0019]

[0020] Taking the elastic model of each element in the finite element model as the adjustment parameter X=(x1, x2, …, x d ), the initial number of fireflies is N, and the Euclidean distance between two fireflies i and j is:

[0021]

[0022] The attraction factor between i and j is:

[0023]

[0024] β0=1, and γ is the attenuation coefficient of light in air, which is 0.01-100;

[0025] When the fitness value of Xj is better than that of Xi, the value of firefly Xi is updated as:

[0026]

[0027] ɑ is a step factor, which is 0-1, and ε is a d-dimensional random variable, which is subject to a standard normal distribution;

[0028] If the fitness value of Xi is the current maximum, Xi will be randomly moved as:

[0029]

[0030] Until the required precision condition is met, the position of the optimal firefly and the luminous intensity of the firefly are output, and the adaptive correction of the digital twin model parameter is completed.

[0031] The application discloses a digital-twin-based offshore wind farm operation and maintenance system, which comprises:

[0032] An offshore wind farm operation data acquisition unit acquires operation data of the offshore wind farm.

[0033] A multi-source heterogeneous data fusion unit performs multi-source heterogeneous data fusion processing on the acquired operation data by using an edge computing method.

[0034] An offshore wind farm digital twin model construction unit constructs a digital twin model of the offshore wind farm by using the operation data subjected to the multi-source heterogeneous data fusion processing.

[0035] A data transmission unit transmits data collected in a physical space to the digital twin model.

[0036] A calculation and analysis unit drives each functional model in the digital twin model to perform calculation and analysis.

[0037] A result feedback unit feeds back calculation and analysis results to the physical space to guide operation and maintenance of the offshore wind farm.

[0038] Preferably, the data transmission unit is arranged in an offshore booster station, accesses an onshore booster station through a fiber ring network, and finally accesses the digital twin model through a TCP / IP protocol.

[0039] Compared with the prior art, the application has the following beneficial technical effects:

[0040] The digital-twin-based offshore wind farm operation and maintenance method disclosed by the application collects signals required for establishing a digital twin system at a data perception level, realizes data fusion through edge computing, establishes a digital twin model driven by mechanism data, activates the model by using signals collected in a physical space, realizes functions such as power prediction, state evaluation, fault early warning, fault diagnosis and life prediction based on the digital twin, and can realize adaptive correction of the digital twin model as the operation state of equipment changes. Through actual application of the digital twin system, the application can effectively reduce field personnel, reduce maintenance time, reduce annual downtime, and reduce unit fault rate, and can serve safe, efficient and intelligent development and utilization of large-scale offshore wind power, and effectively improve the technical level of offshore wind farm operation and maintenance.

[0041] The digital-twin-based offshore wind farm operation and maintenance system disclosed by the application is simple to construct and can be well matched with an existing offshore wind farm operation and maintenance system. BRIEF DESCRIPTION OF DRAWINGS

[0042] Figure 1 The digital twin framework diagram of the present application;

[0043] Figure 2 The physical perception data layer distribution diagram of the offshore wind turbine;

[0044] Figure 3 The digital twin model correction method flow diagram;

[0045] Figure 4 The interface diagram of the offshore wind farm digital twin system. DETAILED DESCRIPTION

[0046] The present application will be further described in conjunction with the drawings and specific embodiments, which are an explanation of the present application rather than a limitation.

[0047] Figure 1 The digital twin framework diagram of the present application, the offshore wind farm operation and maintenance method based on digital twin, includes the following steps:

[0048] Step 1: First, through integrated monitoring technology, realize the collection of key equipment monitoring data of offshore wind farm, provide data support for the functions to be realized by offshore wind digital twin, build physical space, the data needed to be collected include wind turbine blade vibration, blade load, blade video signal, blade sound signal, transmission chain vibration signal, tower vibration, tower load, and wind turbine SCADA data; as shown in Figure 2 ;

[0049] Step 2: Build edge computing module to realize multi-source heterogeneous data fusion, so as to realize the rapid transmission of data; because different data sampling frequencies are different, data types are different, and data volume is huge, edge computing needs to be used to realize data fusion; specifically, for the same type of data, fusion is carried out in the data layer, for example, multiple vibration sensors at the same position of the blade, the method of taking average value is used to realize data fusion; for different types of data, fusion is carried out in the feature layer, wind turbine transmission chain vibration monitoring, through order resampling, Fourier analysis to extract vibration amplitude under different orders, only the data under typical order needs to be saved, so as to greatly reduce the data transmission volume;

[0050] Step 3: The sensor data collection system is arranged in the offshore booster station, the data after data fusion is connected to the onshore booster station through the fiber ring network, and finally connected to the digital twin system through TCP / IP protocol;

[0051] Step 4: Constructing the digital twin model of the offshore wind farm, according to the specific functions realized by the digital twin system, using a knowledge-data dual-driven method to establish a visual model of the offshore wind farm, a wind turbine control model, a wind turbine dynamics model, a wind turbine fault prediction model, a key component life assessment model, an offshore booster station electrical fault diagnosis model, and a submarine cable monitoring and early warning model, ultimately constructing the digital twin of the offshore wind farm;

[0052] Step 5: Passing the physical space collected data to the digital twin model to drive the function models in the twin;

[0053] Step 6: Feedback the digital twin results to the physical space to guide the wind turbine operation and maintenance, including but not limited to comparing the theoretical speed and power of the wind turbine calculated according to the digital twin aerodynamics model with the actual speed and power values in the physical space to evaluate the operation state of the unit; according to the digital twin control model, the best yaw and pitch angles of the unit are simulated in real time and passed to the physical space to realize optimal control based on the digital twin; the digital twin model based on data-driven predicts the oil temperature of the gearbox to realize fault warning of the unit; using the digital twin dynamics model, the load condition of the key components is calculated in real time to realize safety evaluation of the unit, and the fatigue cumulative value is calculated to realize life prediction based on the digital twin;

[0054] As Figure 3 In actual operation, the digital twin model can be self-adaptively corrected, the model parameters are corrected through the real-time interaction characteristics of the digital twin and the physical entity, the iteration update of the twin is realized, the vibration sensors are arranged in 8 layers on the tower of the wind turbine, the 7th order vibration frequency and the corresponding mode shape of the tower can be measured, the actual value and the finite element calculation value of the natural frequency and the mode shape are taken as the objective function, that is:

[0055]

[0056] The elastic model of each element in the finite element model is taken as the adjustment parameter X=(x1,x2,…,x d ), the initial number of fireflies is N, and the Euclidean distance between two fireflies i and j is:

[0057]

[0058] The attraction factor between i and j is:

[0059]

[0060] β0=1, and γ is the attenuation coefficient of light in air, 0.01-100;

[0061] When the fitness value of Xj is better than Xi, the value of firefly Xi is updated as:

[0062]

[0063] ɑ is a step factor, 0~1, and ε is a d-dimensional random variable, obeying a standard normal distribution;

[0064] If the fitness value of Xi is the current maximum, Xi will be randomly moved:

[0065]

[0066] The position of the optimal firefly and the luminous intensity of the firefly are output, and the adaptive correction of the digital twin model parameters is completed.

[0067] As Figure 4 As shown in FIG. 1, it is an interface diagram of the offshore wind farm digital twin system, and the digital twin system based on the wind turbine can realize unit state monitoring, including wind resource information such as wind speed and wind direction, electrical quantities such as power, current and voltage, etc.

[0068] It should be noted that the above only describes some of the embodiments of the present application, and equivalent changes made to the system described in the present application are included in the protection scope of the present application. Those skilled in the art can make similar substitutions to the specific examples described, as long as they do not deviate from the structure of the present application or exceed the scope defined by the present claims, and they are within the protection scope of the present application.

Claims

1. A method for the operation and maintenance of offshore wind farms based on digital twins, characterized in that, include: S1: Collect operational data from offshore wind farms; S2: Using edge computing methods, multi-source heterogeneous data fusion processing is performed on the running data collected by S1; S3: Construct a digital twin model of an offshore wind farm using operational data fusion processing of multi-source heterogeneous data from S2; S4: Transmits the data collected in the physical space to the digital twin model constructed by S3, driving the various functional models in the digital twin model to perform calculations and analyses; S5: Feed back the calculation and analysis results of S4 to the physical space to guide the operation and maintenance of offshore wind farms; In S1, the collected operational data includes blade vibration signals, blade load signals, blade video signals, blade sound signals, transmission chain vibration signals, tower vibration signals, tower load signals, and SCADA data. In S2, data of the same type is fused at the data layer; data of different types is fused at the feature layer. In S3, the functional models in the digital twin model of the offshore wind farm are constructed based on the knowledge-data dual-driven approach. The functional models include a visualization model of offshore wind farms, a control model of wind turbines, a dynamic model of wind turbines, a fault prediction model of wind turbines, a life assessment model of key components, an electrical fault diagnosis model of offshore substations, and a monitoring and early warning model of submarine cables. The constructed digital twin model of the offshore wind farm adaptively corrects the parameters of the digital twin model through the real-time interaction between the digital twin and the physical entity, and achieves iterative updates. The adaptive correction specifically involves arranging vibration sensors in eight layers along the height of the tower to measure the tower's seventh-order vibration frequencies and corresponding mode shapes. The objective function is to minimize the difference between the actual values ​​of the natural frequencies and mode shapes and the finite element calculation values. The elastic model of each element in the finite element model is used as the adjustment parameter. X =(x1,x2,…,x d The initial number of fireflies is N, and the Euclidean distance between two fireflies i and j is: The attraction factor between i and j is: γ is the light attenuation coefficient in air, ranging from 0.01 to 100; When the fitness value of Xj is better than that of Xi, the value of firefly Xi is updated: α is the step size factor, which is 0 to 1, and ε is a d-dimensional random variable that follows a standard normal distribution. If Xi's fitness value is the current maximum, Xi will move randomly: Until the required accuracy conditions are met, the optimal firefly position and luminous intensity are output, completing the adaptive correction of the digital twin model parameters.

2. The offshore wind farm operation and maintenance method based on digital twin as described in claim 1, characterized in that, In S1, operational data of offshore wind farms are collected through integrated monitoring technology.

3. A digital twin-based offshore wind farm operation and maintenance system, used to implement the digital twin-based offshore wind farm operation and maintenance method described in claim 1, characterized in that, include: Offshore wind farm operation data acquisition unit, which collects operation data of offshore wind farms; The multi-source heterogeneous data fusion unit uses edge computing methods to perform multi-source heterogeneous data fusion processing on the collected operational data; The offshore wind farm digital twin model building unit utilizes operational data fusion processing of multi-source heterogeneous data to construct a digital twin model of the offshore wind farm; The data transmission unit transmits data collected in the physical space to the digital twin model; The computational analysis unit drives the various functional models in the digital twin model to perform computational analysis; The results feedback unit feeds back the calculation and analysis results to the physical space to guide the operation and maintenance of offshore wind farms.

4. The offshore wind farm operation and maintenance system based on digital twin as described in claim 3, characterized in that, The data transmission unit is located at the offshore booster station, connected to the onshore booster station via a fiber optic ring network, and finally connected to the digital twin model via the TCP / IP protocol.

Images