Background
The medium voltage direct current distribution network (MVDC) has the advantages of high power generation efficiency, high power distribution efficiency, high reliability and the like, and becomes an important development direction of a future ship power system along with the continuous development of power electronic technology and the increase of the capacity of the ship power system. But the further development of the MVDC of the ship is limited due to the lack of a mature and effective fault protection means. At present, the transient analysis and isolation scheme of a medium-voltage direct-current power distribution network of a high-power ship is relatively lack of design, and especially a mature and effective fault protection means is lack of design, so that the further development of the MVDC of the ship is limited.
In a medium-voltage direct-current power grid of a ship, the condition that a direct-current side low-impedance short-circuit fault occurs in fault current needs to be considered. At this time, the current is rapidly increased and zero-crossing points do not exist, a fault region needs to be rapidly cut off, and a redundant design of fault isolation is considered to ensure the safe operation of medium-voltage direct current of the ship. Aiming at the problem that the direct current side fault current of the MVDC system is difficult to cut off, the MMC with the characteristics of current limiting function, quick restart and the like is one of more ideal choices, and the requirement on a direct current breaker can be reduced. But at present, the analysis and research on the direct-current side fault characteristics of the MMC-MVDC of the ship are less, and a reasonable and rapid protection method is lacked.
In the prior art, due to the fact that the research on the MMC-MVDC protection method of the ship is too few, the experience can be used for reference from the protection method provided by the existing land direct current power grid. Aiming at a VSC-HVDC direct-current transmission system with a large parallel capacitor at a direct-current side, the difference of transient currents of a capacitor branch and a line inlet is measured by using a Pearson correlation coefficient, and the identification of faults inside and outside a zone is realized; however, as the capacitors in the MMC are dispersed in the sub-modules, when the direct current side fails, the MMC sub-modules are in a continuous switching state, and the ship MMC-MVDC cannot perform fault identification according to the method. In the aspect of power distribution network distance measurement, the time-frequency matrix similarity of each branch sample database is analyzed by adopting a full-waveform current characteristic matching technology, so that fault branch positioning is realized; however, the method is mainly used for solving the problem of fault location of the alternating current power distribution network, and a large amount of data needs to be mined and analyzed, so that the requirement of rapidity of MMC-MVDC cannot be met. In the existing ship MVDC protection technology, a method for analyzing transient high-frequency energy by using a frequency domain can solve the problem of selection of thresholds of various running states, but the method has high sampling requirement, the calculation precision is influenced by overlarge frequency domain resolution, and the processing time is long. Therefore, a fault rapid protection method suitable for the marine MMC-MVDC is needed to be found.
Disclosure of Invention
The invention aims to provide a transient current characteristic-based ship MMC-MVDC rapid fault protection method which is mainly used for rapid protection of short-circuit faults and ground faults on a direct-current side circuit.
The purpose of the invention is realized as follows:
step 1, analyzing ship MMC-MVDC fault characteristics, wherein the analysis comprises the analysis of fault current characteristics of an MMC at the moment of a fault and the analysis of fault current characteristics under a multi-region power distribution topological structure, so that an equivalent MMC fault current expression and the relationship between a fault line and a healthy line in a ship region power distribution network are obtained;
step 2: according to the conclusion in the step 1, the characteristics of the influence factors of the fault current are analyzed by combining a correlation algorithm, and a sample database of the fault current information is established;
and step 3: after the fault starting criterion is met, transient current information is sampled, mutation point detection is carried out on the sampled current information, and abnormal values are eliminated; calculating the correlation degree of the sampled current information and the fault current information in the sample database by adopting an improved cosine similarity algorithm to complete fault detection, and presuming the approximate fault transition resistance range according to the fault current information with the maximum correlation degree in the sample database;
and 4, step 4: replacing the current direction at the moment of the fault according to the positivity and the negativity of the correlation degree of each line meeting the fault threshold; and further, the fault location is more reliably and accurately realized aiming at the characteristics of current directions of different fault areas.
Compared with the prior art, the invention has the beneficial effects that:
the method theoretically analyzes the bipolar short-circuit fault characteristics of the medium-voltage direct-current power distribution network of the ship, and establishes a local database of fault current information from the aspect of fault current influence factors; meanwhile, mutation point detection is carried out to ensure the accuracy of related analysis results; the fault detection method provided by the invention has the advantages of high action speed, good selectivity, low communication requirement and capability of resisting transition resistance, and can solve the defects of long fault detection time, complex detection algorithm and the like of the MMC-MVDC of the existing ship.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
The technical solutions in the embodiments of the present invention are clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
Embodiments of the invention will be described in further detail below with reference to the accompanying drawings, the method comprising:
step 1, analyzing the fault characteristics of the MMC-MVDC of the ship, which is mainly divided into two steps
The first step is as follows: the characteristic analysis of fault current at the moment of short-circuit fault of the direct current side of MMC in the MMC-MVDC is carried out by referring to an equivalent RLC circuit of FIG. 2, wherein the expression of the fault current is
Wherein the root of the characteristic equation is
α,ω
0Respectively representing an impedance factor and a resonance frequency, which are respectively alpha-R/2L,
U
dc(0)、i
dc(0) dividing the responses of the voltage value and the current value at the initial time of the fault into under damping, critical damping and over damping which respectively correspond to the under damping, the critical damping and the over damping
The situation is.
1.1) under-damped condition, the capacitor discharge current at fault can be expressed as
The time at which the current peaks can be approximately calculated:
2.2) over-damped condition
The time at which the current peaks can be approximately calculated:
the second step is that: due to the fact that the MVDC circuit of the ship is short, the parameters of the RL model are far smaller than those of the MMC and the line reactor, and the transient fault current characteristics of the MMC can hardly be influenced by the line.
Specific details are set forth in FIG. 3 in accordance with an embodiment of the present invention. In fig. 3, for the annularly connected regional distribution network, the MMC not connected to the bus where the fault line is located is selected to decouple and open the ring network, and the most serious line fault component current is calculated. At this time, the fault line current I10And healthy line current I31The relationship is as follows:
relationship between healthy line and faulty line:
wherein k is I31/I32。
Step 2: according to the conclusion in the step 1, the characteristics of the influence factors of the fault current are analyzed by combining a correlation algorithm, and a fault current information sample database is established;
at the moment of fault occurrence, the direct current transient current is mainly affected by fault transition resistance and fault initial current according to a fault current expression.
In particular, the current characteristics influenced by each parameter can be embodied by changing the slope.
At the moment of failure occurrence, assume time t (0)+) The magnitude of the current slope can be expressed as
From the above formula, it can be seen that the current slope at the moment of the fault does not change much, and the fault current increases approximately linearly from the initial current; but as the resistance increases, the current slope changes significantly. When the fault transition resistance is small, the linear characteristic of the change rate of the fault current is hardly influenced by the initial fault current, and only the amplitude of the fault current is influenced. The larger the fault initiation current, the more pronounced the change in current slope as the fault transition resistance increases.
Firstly, the established fault current sample database needs to consider the characteristics of the influence of fault transition resistance and fault initial current on fault current, and provides necessary data for accurate judgment.
Wherein r is1~rnRepresenting different fault resistances, I0~ImTransient current data representing different fault initiation currents.
Secondly, analyzing the established sample database by adopting a cosine similarity algorithm, verifying the characteristics of fault influence factors and reducing unnecessary data.
In the formula:<Yi,Yj>to representThe inner product of matrices Y (i, n) and Y (j, n); y is the norm of matrix Y.
Analogous to the vector angle, γ is defined as the angle between 2 matrices: when γ is 90, p is 0, meaning that 2 matrices are completely different; when γ is 0, p is 1, indicating that 2 matrices have extremely high similarity.
And step 3: after the fault starting criterion is met, transient current information is sampled, mutation point detection is carried out on the sampled current information, abnormal values are eliminated, the correlation degree is obtained by adopting an improved cosine similarity algorithm, fault detection is completed, and the approximate fault transition resistance range is presumed according to the correlation information on the outgoing line side of the converter;
the fault starting criterion is mature at present, and mainly comprises starting criteria such as under-voltage starting, over-current starting, gradient current starting, current change rate and voltage change rate.
In order to eliminate abnormal values, the sampled data needs to be preliminarily screened, so that the influence on the overall correlation degree due to the abnormal occurrence of a certain single data is prevented;
in the sampled data, according to (t, I (t)) as a two-dimensional array, a first sampling point is used as an origin, the cosine similarity of each point is calculated, if the cosine value is smaller than a selected threshold value, abnormal points are removed, if more abnormal values exist and the cosine value is in a reasonable range, a second sampling point is used as the origin, the cosine similarity of each point behind is calculated until the abnormal points are smaller than a certain range.
In the formula:<A1,A2>representation matrix A1And A2Inner product of (d); | | A1I is the norm of the matrix A, x0,xi,xi+1Is a specific value in the sampled current.
And (3) combining the transient current analysis in the step (2), and under the condition that the fault transition resistance is small, the influence of the fault initial current on the current slope is small. Therefore, the sampling current data with the abnormal fault points removed is adjusted by adopting the improved cosine similarity and then is compared with the current data of the fault sample database. For the case of large fault transition resistance, the establishment of the initial current sample for the fault needs to be considered.
And for the two ends of the line with the detected fault position, the positive and negative values of the correlation indicate the direction of the fault current to carry out accurate fault positioning.
And 4, step 4: the fault location of different areas is completed aiming at the characteristics of different fault areas and the signs of fault correlation, the direction of an incoming line is taken as positive, and the location logic of each fault is as follows:
for the fault F2 located in the power generation section: rho > 0&&|ρ|>δG
For faults F3, F4 located in the load section: rho > 0&&|ρ|>δload
Fault location criteria for line sections:
the left end of the circuit: rhol>0&&|ρl|>δline(ii) a The right end of the circuit: rhor>0&&|ρr|>δline。
Meanwhile, the line fault can be accurately positioned.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.