An environmental identification method for precise synchronization of power systems

Abstract

This invention relates to an environment identification method for precise synchronization of power systems, primarily addressing the problem of large delay estimation errors in existing technologies under non-line-of-sight (LAS) transmission channels. The implementation scheme involves dividing the existing power grid channel environment into LAS and LAS channels. Feature parameters are extracted from the datasets of both channels under the power grid environment using wavelet transform. The feature values ​​are then trained using an input vector machine (IVM) to construct a classifier model. Based on the identified channel type, for LAS channels, an incoherent estimator is used to estimate the propagation delay; for LAS channels, a signal reconstruction method is used to obtain the first path signal, thereby eliminating its interference with delay estimation and obtaining an accurate delay estimate. Finally, the power equipment time is adjusted based on the estimate to ensure data synchronization. This invention can estimate delays under both LAS and LAS channels and can be used for precise time synchronization of power equipment in complex communication environments.

Description

Technical Field

[0001] This invention belongs to the field of power grid communication technology, and specifically relates to an environment identification method for precise synchronization of power systems. It can be used for precise time synchronization of power systems in complex channel environments, and realizes synchronous exchange of data packets between various power devices. Background Technology

[0002] As power grid services become increasingly diversified, wireless communication needs to be integrated with traditional power systems to address service access challenges. Examples include robotic inspections, on-site maintenance and collaborative commissioning of power equipment, and mobile inspections of substations. These applications place more complex demands on time synchronization between various power devices.

[0003] Time synchronization between devices can be achieved by estimating propagation delay. In line-of-sight (LOS) channels, multipath amplitudes follow a Ricean distribution, and the intensity of the direct path component is much greater than that of the other paths, allowing the generalized cross-correlation algorithm to estimate the delay. For non-line-of-sight (NLOS) channels, multipath amplitudes follow a Rayleigh distribution, with smaller differences in the intensity of each path, causing greater interference to delay estimation. Therefore, it is necessary to reconstruct the received signal before performing delay estimation.

[0004] The communication environment in power grid business scenarios is generally complex, so it is necessary to identify the specific channel environment in order to obtain accurate time delay. Based on this, machine learning can be used to classify and identify the channel environment. Input Vector Machine (IVM), as a machine learning algorithm, is widely used in classification problems and has the following significant advantages: (1) it achieves the highest classification accuracy by means of feature selection strategy; (2) it achieves minimum sparsity using very few input vectors; and (3) it provides probabilistic interpretation of the classification results. Therefore, this technology can be chosen to identify the channel environment. This enables precise time synchronization of the power system in complex environments and ensures the correct transmission of data packets.

[0005] Patent ZL201210084420.3 discloses a channel environment detection method using Support Vector Machine (SVM). This method takes the change in signal estimation and the amplitude of data block transitions as two channel attributes, and whether the channel environment has changed as the initial values ​​for the SVM. The K-CV method is used to determine the penalty parameter c and the function parameter g, and the radial basis function kernel function is used as the SVM parameters to build the SVM model and detect whether the channel environment has changed. This method ensures the accuracy of channel environment detection even when the channel environment is completely unknown. However, it can only detect whether the channel environment has changed; it cannot identify specific channel environments.

[0006] Patent ZL 201810096220.7 discloses a multipath delay estimation method and apparatus. This method utilizes DMRS to obtain channel estimates in the frequency domain corresponding to each subcarrier, then performs a time-domain transformation on the denoised channel estimates in the frequency domain, and finally performs a time-domain multipath search to determine the multipath delay in the OFDM symbol carrying the DMRS. This method can filter out non-multipath noise and interference terms during the multipath delay estimation process, improving the accuracy of multipath delay estimation. However, its performance degrades in the presence of inter-cell interference, and it is also prone to search errors in NLOS environments, leading to increased delay estimation errors. Summary of the Invention

[0007] The purpose of this invention is to address the shortcomings of the existing technology by providing an environmental identification method for precise synchronization of power systems. This method uses machine learning to classify and identify power grid channels and processes them accordingly based on the identified channel types, thereby ensuring time synchronization of various communication power devices in complex environments and achieving consistent information and data exchange.

[0008] To achieve the above objectives, the technical solution of the present invention includes the following:

[0009] 1) Establish a 5G power grid system model. The 5G smart grid consists of 5G base stations and power equipment, including power-consuming equipment and power distribution equipment, all of which integrate 5G communication modules;

[0010] 2) Base stations serving 5G smart grids classify the existing power grid channel environment and construct a channel identification model under the power system:

[0011] 2a) Divide the existing power grid channel environment into line-of-sight (LOS) transmission and non-line-of-sight (LOS) transmission.

[0012] NLOS channel;

[0013] 2b) Obtain typical channel data under power grid conditions, and process it to obtain datasets of LOS and NLOS channels;

[0014] 2c) Extract channel time-frequency characteristic parameters from the dataset using wavelet transform;

[0015] 2d) Train the model on the dataset based on the Input Vector Machine (IVM);

[0016] 2e) Optimize the parameters of the constructed channel identification model;

[0017] 3) Power equipment sends existing pilot signals to 5G base stations;

[0018] 4) 5G base stations identify the channel environment and select the appropriate processing method:

[0019] If the current channel environment is an NLOS channel, proceed to step 5);

[0020] If the current channel environment is a LOS channel, proceed to step 6);

[0021] 5) 5G base stations reconstruct received signals to cancel out non-first-path signals:

[0022] 5a) Given the pilot signal S T With received signal S R Cross-correlation yields R A (τ), the formula is as follows:

[0023]

[0024] Among them, K A The known pilot signal S T With received signal S R The length of the cross-correlation results;

[0025] 5b) According to R A (τ) The formula for calculating the amplitude loss Ami of the i-th non-first-path signal component is as follows:

[0026] 5c) Calculate the non-first-path signal S based on amplitude loss. NLOS The formula is as follows:

[0027]

[0028] in p M is the multipath number;

[0029] 5d) Based on the received signal S R Non-first-path signal S NLOS Obtain the first diameter signal S NC The formula is as follows:

[0030] S NC =S R -S NLOS ;

[0031] 6) The 5G base station estimates the propagation delay i based on the first path signal and sends it to the power equipment;

[0032] 6a) Given the pilot signal S T With the first diameter signal S NC Cross-correlation yields R(τ);

[0033] 6b) The propagation delay i is estimated from R(τ) using an incoherent estimator and sent to the power equipment;

[0034] 7) Root propagation delay of power equipment Adjust existing time information to achieve data synchronization and exchange.

[0035] Compared with the prior art, the present invention has the following advantages:

[0036] 1) Existing technologies only detect changes in the channel environment and cannot specifically identify the channel type. This invention extracts feature parameters from the power grid communication environment using wavelet transform, trains the feature values ​​based on input vector machine (IVM) to construct a classifier model, and optimizes it using a combination of network search and cross-validation. This enables the model to determine the channel type in complex environments, providing a foundation for delay estimation.

[0037] 2) Existing technologies only estimate propagation delay for LOS channels, resulting in significant errors for NLOS channels and making them unsuitable for complex power grid environments. This invention identifies channel types through machine learning. For LOS channels, an incoherent estimator is used to estimate propagation delay; for NLOS channels, a signal reconstruction method is used to obtain the first path signal, thereby eliminating interference from the NLOS channel on delay estimation and obtaining an accurate delay estimate. This achieves channel identification in complex environments, ensuring accurate time synchronization of the power system. It is applicable to both LOS and NLOS channels. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the overall implementation process of the present invention;

[0039] Figure 2 This is a block diagram of a power grid system under LOS channel in this invention;

[0040] Figure 3 This is a block diagram of a power grid system under NLOS channel in this invention;

[0041] Figure 4 This is a flowchart of the working mode switching sub-process in this invention;

[0042] Figure 5 This is a diagram of the detection signal frame structure in this invention.

[0043] 1. Building, 2. Power supply dispatch center, 3. Base station, 4. Power distribution equipment, 5. Non-MBSFN area / control area, 6. MVSFN area, A represents time slot 1, B represents time slot 2. Detailed Implementation

[0044] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0045] like Figure 1 This is a schematic diagram of the overall implementation process of the present invention. The present invention adopts power grid systems under both LOS and NLOS channel environments, such as... Figure 2This is a block diagram of a power grid system under LOS channel. Figure 3 This is a block diagram of a power grid system under NLOS channel. The electrical equipment and power distribution equipment 4 in building 1 are users who communicate with base station 3 in the power grid communication environment, and both are equipped with 5G communication modules.

[0046] The present invention provides a method for ensuring time synchronization of power equipment under LOS and NLOS channels, the implementation steps of which are as follows:

[0047] Step 1: Establish a 5G power grid system model. The 5G smart grid consists of 5G base stations 3 and power equipment 4. The equipment includes power-consuming equipment and power-distributing equipment, all of which integrate 5G communication modules.

[0048] Step 2: Base station 3 serving the 5G smart grid classifies the existing power grid channel environment and constructs a channel identification model under the power system;

[0049] The specific implementation method for classifying the existing power grid channel environment by base stations serving 5G smart grids is as follows:

[0050] 2.1) The existing power grid channel environment is divided into line-of-sight (LOS) and non-line-of-sight (NLOS) channels.

[0051] 2.2) Obtain typical channel data under the power grid environment and process it to obtain datasets of LOS and NLOS channels.

[0052] 2.3) Select the complex Morlet wavelet as the mother wavelet and use wavelet transform to extract the channel time-frequency characteristic parameters from the dataset.

[0053] 2.4) Train the model on the dataset based on the input vector machine (IVM).

[0054] 2.5) The parameters of the constructed channel identification model are tuned by combining network search and cross-validation.

[0055] Step 3: Power equipment 4 sends a known pilot signal to 5G base station 3. The pilot signal uses a Synchronous Broadcast Block (SSB), referencing... Figure 4 This is a flowchart for switching work modes.

[0056] In the diagram, area 6 represents the pilot SSB required for precise synchronization within the MVSFN region, while area 5 is a non-MBSFN / control area and cannot be used to carry SSBs. The placement of the SSB in area 6 is optional depending on the subcarrier and frequency band; the diagram shows an example of SSB Case A. The default SSB transmission period is 20ms, which can be changed according to actual needs, with optional configurations of 5ms, 10ms, 20ms, 40ms, 80ms, and 160ms.

[0057] Step 4: 5G base station 3 identifies the channel environment and selects the processing method:

[0058] If the current channel environment is an NLOS channel, proceed to step 5;

[0059] If the current channel environment is a LOS channel, proceed to step 6.

[0060] Step 5: 5G reconstructs the received signal to cancel out non-first-path signals, such as... Figure 5 The diagram shown is a frame structure diagram of the detection signal.

[0061] The operation and calculation steps of step S5 include the following:

[0062] 1) Given the pilot signal S T With received signal S R Cross-correlation yields R A (τ), the calculation formula is as follows:

[0063]

[0064] 2) According to R A (τ) Calculate the amplitude loss Ami of the i-th non-first-path signal component, as follows:

[0065]

[0066] 3) Calculate the non-first-path signal S based on amplitude loss. NLOS The formula is as follows:

[0067]

[0068] in p M is the multipath number;

[0069] 4) Based on the received signal S R Non-first-path signal S NLOS Obtain the first diameter signal S NC The formula is as follows:

[0070] S NC =S R -S NLOS .

[0071] Step 6: The 5G base station estimates the propagation delay based on the first-path signal. And send it to the power equipment.

[0072] The specific implementation method for this step is as follows:

[0073] 6.1) Calculate the known pilot signal S T With the first diameter signal S NC The cross-correlation R(τ) is given by the following formula:

[0074]

[0075] Where M is the length of the pilot signal and W is the length of the relevant time window.

[0076] 6.2) The propagation delay is estimated from R(τ) using an incoherent estimator. Users based on estimated propagation delay To adjust existing time information, use the following formula:

[0077]

[0078] Step 7, Root Propagation Delay of Power Equipment Adjust existing time information to achieve data synchronization and exchange.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them; although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications can still be made to the specific implementation of the present invention or equivalent substitutions can be made to some technical features without departing from the spirit of the technical solutions of the present invention, and all such modifications and substitutions should be covered within the scope of the technical solutions claimed in the present invention.

Claims

1. An environmental identification method for precise synchronization of power systems, characterized in that, Includes the following steps: S1. Establish a 5G power grid system model; S2. Classify the existing power grid channel environment and construct a channel identification model under the power system; S3. Power equipment sends known pilot signals to 5G base stations, and the pilot signals use Synchronous Broadcast Block (SSB). S4. The base station identifies the channel environment and selects the appropriate processing method: If the current channel environment is an NLOS channel, proceed to step S5); If the current channel environment is a LOS channel, proceed to step S6); S5.5G base stations reconstruct received signals to cancel out non-first-path signals; S6.5G base stations estimate propagation delay based on the first-path signal. And send it to the electrical equipment: S7. Power equipment based on propagation delay Adjust existing time information to achieve data synchronization and exchange; The operation steps of step S5 include the following: 1) Given the pilot signal S T With received signal S R Cross-correlation yields R A (τ), the calculation formula is as follows: 2) According to R A (τ) Calculate the amplitude loss Am of the i-th non-first-path signal component. i The calculation formula is as follows: 3) Calculate the non-first-path signal S based on amplitude loss. NLOS The calculation formula is as follows: Among them, M p It is a multipath number; 4) Based on the received signal S R Non-first-path signal S NLOS Obtain the first diameter signal S NC The calculation formula is as follows: S NC =S R -S NLOS 。 2. The environmental identification method for precise synchronization of power systems according to claim 1, characterized in that, The method for classifying the existing power grid channel environment in step S2 includes: a) Divide the existing power grid channel environment into line-of-sight (LOS) and non-line-of-sight (NLOS) channels; b) Obtain typical channel data under power grid conditions and process it to obtain datasets of LOS and NLOS channels; c) Extract channel time-frequency characteristic parameters from the dataset using wavelet transform; d) Train a channel identification model for the power system on the dataset based on the input vector machine (IVM); e) Optimize the parameters of the constructed channel identification model.

3. The environmental identification method for precise synchronization of power systems according to claim 1, characterized in that, The operation steps of step S6 include the following: 1) Given the pilot signal S T With the first diameter signal S NC Cross-correlation yields R(τ); 2) The propagation delay is estimated from R(τ) using an incoherent estimator. And send it to the power equipment.

4. The environmental identification method for precise synchronization of power systems according to claim 2, characterized in that, Step d) involves using IVM to classify and identify LOS and NLOS channels, specifically including: The IVM model and IVM classifier are obtained by training feature data; Based on the trained IVM classifier, determine the LOS channel and NLOS channel.

5. The environmental identification method for precise synchronization of power systems according to claim 1, characterized in that, In step S6, the propagation delay is estimated based on the first path signal. The implementation is as follows: 6a) Calculate the known pilot signal S T With the first diameter signal S NC Cross-correlation R(τ): Where M is the length of the pilot signal, and W is the length of the relevant time window. 6b) Calculate propagation delay 6. The environmental identification method for precise synchronization of power systems according to claim 2, characterized in that, The classification method described in step a) first divides the power grid channel environment into line-of-sight (LOS) channels and non-line-of-sight (NLOS) channels. In the LOS channel, the strongest path can be used as the first path for direct time delay estimation. In the NLOS channel, the signal needs to be reconstructed to obtain the first path signal.

7. The environmental identification method for precise synchronization of power systems according to claim 6, characterized in that, The channel environment of the LOS channel and the non-line-of-sight (NLOS) transmission channel is a time-frequency channel dataset composed of channel environment parameters.

8. The environmental identification method for precise synchronization of power systems according to claim 7, characterized in that, In step 3), the channel environment parameters are extracted by wavelet transform. The complex Morlet wavelet is selected as the mother wavelet. Through multiple wavelet transforms, the wavelet power spectrum image can be obtained. The wavelet power spectrum is normalized and used as the input of the IVM classifier.

9. The environmental identification method for precise synchronization of power systems according to claim 2, characterized in that, In step e), the parameter tuning of the constructed model adopts a combination of network search and cross-validation.

Images