VHF / UHF radio wave propagation prediction model construction method based on entropy weight fusion

Abstract

The invention discloses a VHF / UHF radio wave propagation prediction model construction method based on entropy weight fusion, and the method comprises the steps: collecting measurement data, carrying out the scaling of a right hyper-parameter, and dividing the measurement data into a training set, a test set and a verification set according to a proportion; iTU-R P.1546 and 2001 models are used as sub-models, and a combined prediction model is established based on an entropy weight theory; and testing, adjusting and verifying the combined prediction model by using the data of the test set and the verification set, so that the precision of the combined prediction model reaches the expectation. According to the method, the frequency band utilization rate of the VHF / UHF frequency band can be improved, so that mutual interference between electronic information systems is reduced, and technical support can be provided for propagation loss prediction of radio signals of the VHF / UHF frequency band, electromagnetic environment analysis, performance evaluation and comprehensive application of the electronic information systems and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of VHF / UHF frequency band radio propagation prediction, and in particular to a method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion. Background Art

[0002] VHF and UHF are the most widely used, most efficient and most crowded frequency bands. They are important frequency bands for current and future radio communications. They are widely used in various businesses such as communications, broadcasting, radar and radio astronomy. As a result, mutual interference between various electronic information systems will undoubtedly reduce the performance of the communication system. Therefore, increasing the capacity of this frequency band and reducing the interference between various electronic information systems have become urgent issues to be solved.

[0003] As the transmission path between the transceiver devices of the electronic information system, the wireless channel is affected by many factors such as topography, troposphere, ionosphere, solar activity, etc., and the channel characteristics determine the performance of the electronic information system. Therefore, the study of high-precision channel models has become an important breakthrough point to alleviate the frequency tension, and improving the prediction accuracy of VHF / UHF band radio wave propagation, improving the frequency band utilization of VHF / UHF band, and thus reducing the mutual interference between electronic information systems has become an urgent problem to be solved in this field. Summary of the invention

[0004] The purpose of the present invention is to improve the prediction accuracy of radio wave propagation in the VHF / UHF frequency band, and to provide a method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion, which is used to improve the frequency band utilization of the VHF / UHF frequency band, thereby reducing mutual interference between electronic information systems, and can provide technical support for the propagation loss prediction of radio signals in the VHF / UHF frequency band, electromagnetic environment analysis, electronic information system performance evaluation and comprehensive application.

[0005] A method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion, comprising:

[0006] S1. Collect measurement data, scale the hyperparameters for weight confirmation, and then divide the measurement data into training set, test set, and validation set in proportion;

[0007] S2. Taking ITU-R P.1546 and 2001 models as sub-models, a combined prediction model is established based on entropy weight theory;

[0008] S3. Use the test set and validation set data to test, adjust and validate the combined prediction model so that the accuracy of the combined prediction model reaches the expected level.

[0009] In step S1, the collected measurement data includes the transmitting antenna height h1, the receiving antenna height h2, the longitude and latitude position of the transmitting antenna (L o1 ,L a1 ), the longitude and latitude position of the receiving antenna (L o2 ,L a2 ), antenna transmission angle θ, signal frequency f and terrain clearance angle

[0010] Among them, in step S1, after collecting the measurement data and before dividing the data set, the measured values ​​of the confirmation hyperparameters and the sub-model prediction values ​​are scaled by normalization so that the measured values ​​of the confirmation hyperparameters and the sub-model prediction values ​​have the same contribution or effect in confirmation.

[0011] The confirmation hyper parameters include signal frequency f, transceiver distance d, transmitting antenna height h1, receiving antenna height h2 and terrain clearance angle

[0012] In step S1, the collected measurement data is divided into a training set, a validation set and a test set in a ratio of 6:2:2.

[0013] In step S2, the ITU-R P.1546 and 2001 models are used as sub-models, and a combined prediction model is established based on the entropy weight theory, including:

[0014] Step 21. Construct the discriminant matrix:

[0015] The values ​​in the discriminant matrix are composed of the absolute values ​​of the difference between the measured values ​​and the predicted values ​​of the sub-models; assuming that each frequency band has n measured values, the absolute value of the difference between the measured value and the predicted value at the j-th training point of the i-th independent prediction model is Lr ij ; Then the judgment matrix V is:

[0016]

[0017] Where, i = {1, 2}, i = 1 corresponds to the ITU-R P.1546 model, i = 2 corresponds to the ITU-R P.2001 model; j = {1, 2, ..., n}, corresponding to n measured values ​​at each frequency;

[0018] Step 22. Calculate the weight p of the sub-model at each training point based on the value in the discriminant matrix ij ;

[0019]

[0020] Step 23. Based on the weight p of the sub-model at each training point ij Calculate the sub-model information entropy H i ;

[0021]

[0022] Step 24. Based on the sub-model information entropy H i Calculate the weight W of each sub-model in the combined prediction model i ;

[0023]

[0024] Step 25. Based on the weight W of each sub-model in the combined prediction model i , establish a combined prediction model, and use the combined prediction model to predict the path loss L C =w1×L1+w2×L2;

[0025] Where, L c is the path loss, L1 is the value predicted by the ITU-R P.1546 model; W1 is the weight of the ITU-R P.1546 model; L2 is the value predicted by the ITU-R P.2001 model; W2 is the weight of the ITU-R P.2001 model.

[0026] Among them, the absolute value of the difference between the measured value and the predicted value at the jth training point of the i-th independent prediction model is Lr ij Obtained by the following formula:

[0027] L ij =|Lr obs -Lr pre |;

[0028] In the formula, Lr obs is the measured value, Lr pre is the predicted value.

[0029] Among them, the sum of the weights of the two sub-models is equal to 1.

[0030] The combined prediction model input is the signal frequency f, the longitude and latitude position of the transmitting antenna (L o1 ,L a1 ), the longitude and latitude position of the receiving antenna (L o2 ,L a2 ), ITU-R P.1546 and 2001 model predicted loss values, and the output is the final path loss L calculated by the combined prediction model c .

[0031] The method of testing, adjusting and verifying the combined prediction model using the test set and validation set data so that the accuracy of the combined prediction model reaches the expected level includes:

[0032] Use the combined prediction model to predict the test set data, and compare and analyze the predicted value with the measured value and the sub-model predicted value to determine whether the prediction accuracy of the combined prediction model meets expectations; if the prediction accuracy of the combined prediction model is lower than that of the prediction sub-model and does not meet expectations, replace the confirmation hyperparameters and repeat step S2;

[0033] When the prediction accuracy of the combined prediction model reaches the expected level, the final prediction model is output, which is the VHF / UHF radio wave propagation combined prediction model based on entropy weight theory.

[0034] The VHF / UHF radio wave propagation combined prediction model based on entropy weight theory constructed by the present invention can improve the prediction accuracy of VHF / UHF frequency band radio wave propagation, thereby reducing interference between wireless information systems and improving transmission quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a construction flow chart of a method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion according to an embodiment of the present invention;

[0036] Figure 2 is a flow chart of modeling data collection and preprocessing according to an embodiment of the present invention;

[0037] Figure 3 is a flow chart of establishing a combined prediction model based on entropy theory according to an embodiment of the present invention;

[0038] Figure 4 It is a flow chart of model testing and adjustment according to an embodiment of the present invention.

[0039] Figure 5 It is a schematic diagram of a VHF / UHF radio wave propagation combined prediction model based on entropy weight theory constructed in an embodiment of the present invention. DETAILED DESCRIPTION

[0040] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0041] The ITU-R P.1546 and 2001 models proposed by the International Telecommunication Union (ITU) have been continuously revised and improved for nearly a decade and are widely used for propagation prediction in the VHF / UHF band. Among them, the ITU-R P.1546 recommendation is applicable to the prediction of point-to-area radio propagation for terrestrial services in the frequency range of 30MHz to 4000MHz. This method can be used for the prediction of tropospheric radio propagation on land paths, sea paths and / or land-sea mixed paths with effective transmitting antenna heights less than 3000m and path lengths between 1 and 1000km. It can interpolate, extrapolate and correct the empirically derived field strength curves to evaluate the impact of terrain clearance and ground obstruction on propagation.

[0042] Recommendation ITU-R P.2001 is a widely used model for terrestrial propagation, covering the frequency range of 30 MHz to 50 GHz and distances from 3 km to 1000 km.

[0043] Entropy weight theory is a modeling method that uses the concept of entropy to determine the amount of information feedback from each attribute or indicator in the system, and assigns each attribute or indicator a corresponding weight value based on the amount of information.

[0044] The inventors of the present invention intend to use the entropy weight theory to fuse the ITU-R P.1546 and 2001 models to establish a high-precision combined prediction model, from the perspective of improving the prediction accuracy of the model, thereby reducing the problem of mutual interference between electronic information systems.

[0045] The method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion in an embodiment of the present invention is implemented by the following technical solutions:

[0046] Step A: Modeling data collection and preprocessing.

[0047] Collect measurement data, scale the hyperparameters, and divide the collected data into training, test, and validation sets in proportion for model building.

[0048] Step B: Establish a combined prediction model based on entropy theory.

[0049] Taking ITU-R P.1546 and 2001 models as sub-models, a combined prediction model is established based on entropy weight theory.

[0050] Step C: Model testing and adjustment.

[0051] The test set and validation set data are used to test, adjust and verify the combined prediction model to achieve the prediction effect.

[0052] Further, step A of the embodiment of the present invention specifically includes:

[0053] Step A1: Modeling data collection.

[0054] The modeling data required by the present invention include the height of the transmitting antenna h1 (unit dimension is m), the height of the receiving antenna h2 (unit dimension is m), the position of the transmitting antenna (L o1 ,L a1 )(L o Indicates longitude, L a indicates latitude), the location of the receiving antenna (L o2 ,L a2 ), antenna transmission angle θ (unit dimension is °), signal frequency f (unit dimension is MHz) and terrain clearance angle (unit dimension is °) and other related parameters.

[0055] Step A2: Confirm (hyper)parameter scaling.

[0056] The embodiment of the present invention calculates the information entropy contained in the sub-model by predicting the relative error or root mean square error between the sub-model and the measured value, thereby determining the weight of the sub-model using the entropy weight method, wherein the weight (hyper) parameters include f, transceiver distance d (unit dimension is km), transmitting antenna height h1, receiving antenna height h2 and terrain clearance angle Since there is a large difference in order of magnitude between the relative error and the root mean square error, the embodiment of the present invention scales the measured value and the ITU-R P.1546 and 2001 model prediction value by normalization, so that they achieve the same contribution in terms of right confirmation.

[0057] Step A3: Dataset division.

[0058] The data collected from typical areas are divided into training set, validation set and test set in a ratio of 6:2:2.

[0059] Further, step B of the embodiment of the present invention specifically includes:

[0060] Step B1: Construct the discriminant matrix.

[0061] The values ​​in the discriminant matrix are composed of the absolute values ​​of the differences between the measured values ​​and the sub-model predicted values.

[0062] Assume that each frequency band has n measured values, and the absolute value of the difference between the measured value and the predicted value at the jth training point of the i-th independent model is Lr ij , the calculation formula is shown in formula (2), then the judgment matrix V is:

[0063]

[0064] Wherein, i = {1, 2}, i = 1 corresponds to the ITU-R P.1546 model, i = 2 corresponds to the ITU-R P.2001 model; j = {1, 2, ..., n}, corresponding to n measured values ​​at each frequency.

[0065] L ij =|Lr obs -Lr pre | (2)

[0066] Where Lr obs is the measured value, Lr pre is the predicted value.

[0067] Step B2: Calculate the weight of the sub-model at the training point.

[0068] Using the constructed discriminant matrix, the weight p of the sub-model at each training point is calculated according to formula (3): ij .

[0069]

[0070] Step B3: Calculate sub-model information entropy.

[0071] After constructing the discriminant matrix, use formula (4) to calculate the information entropy H contained in the sub-model i size.

[0072]

[0073] In the formula, H i is the Shannon entropy of the ith independent model, p ij is the weight of the ith independent prediction model at the jth measured value, and the calculation formula is shown in formula (2).

[0074] Step B4: Calculate the final weight of the sub-model.

[0075] Substitute the sub-model information entropy into formula (5) to calculate the weight W of each sub-model in the combined prediction model i .

[0076]

[0077] Where 0≤w 1,2 ≤1,w1+w2=1.

[0078] Step B5: Establishment of combined prediction model.

[0079] Substitute the weights of each sub-model calculated by formula (5) and the predicted value of the sub-model into formula (6) for weighted calculation to obtain the final transmission loss L c .

[0080] L C =w1×L1+w2×L2 (6)

[0081] Where L1 is the predicted value of the ITU-R P.1546 model; W1 is the calculated ITU-R P.1546 model weight, L2 is the predicted value of the ITU-R P.2001 model; W2 is the calculated ITU-R P.2001 model weight.

[0082] Further, step C of the embodiment of the present invention includes:

[0083] Step C1: Model performance test.

[0084] The constructed combined prediction model is used to predict the test set data and compared with the measured values ​​and sub-model prediction values ​​to determine whether the prediction accuracy of the proposed model meets the requirements.

[0085] Step C2: Change the rights confirmation parameters.

[0086] If the prediction accuracy of the constructed model is lower than that of the prediction sub-model and does not meet the expected requirements, replace the confirmation hyperparameters and repeat step B.

[0087] Step C3: Output the final model.

[0088] If the prediction accuracy of the established model meets the requirements, the final prediction model is output, and a combined prediction model of VHF / UHF radio wave propagation based on entropy weight theory is established.

[0089] The combined prediction model input of the embodiment of the present invention is f, (L o1 ,L a1 )、(L o2 ,L a2 ), ITU-R P.1546 and 2001 model prediction (loss) values ​​L1, L2, the output is the final path loss L calculated by the constructed combined prediction model c , the model prediction diagram is as follows Figure 5 shown.

[0090] Among them, the weights W1 and W2 of the sub-models in the combined prediction model are obtained through the entropy weight fusion model, and the input of the entropy weight fusion model is f, (L o1 ,L a1 )、(L o2 ,L a2 ), the output of the entropy weight fusion model is the weights W1, W2 of the sub-models in the combined prediction model.

[0091] The VHF / UHF radio wave propagation combined prediction model based on entropy weight theory constructed in the embodiment of the present invention can improve the prediction accuracy of VHF / UHF frequency band radio wave propagation, thereby reducing interference between wireless information systems and improving transmission quality.

[0092] The basic principles and main features of the present invention and the advantages of the present invention are shown and described above. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention.

[0093] Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the claims are included in the present invention.

[0094] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion, characterized in that: include: S1. Collect measurement data, scale the hyperparameters for weight confirmation, and then divide the measurement data into training set, test set, and validation set in proportion; S2. Taking ITU-R P.1546 and 2001 models as sub-models, a combined prediction model is established based on entropy weight theory; S3. Use the test set and validation set data to test, adjust and validate the combined prediction model so that the accuracy of the combined prediction model reaches the expected level.

2. The method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion according to claim 1, characterized in that: In step S1, the collected measurement data includes the transmitting antenna height h1, the receiving antenna height h2, the longitude and latitude position of the transmitting antenna (L o1 ,L a1 ), the longitude and latitude position of the receiving antenna (L o2 ,L a2 ), antenna transmission angle θ, signal frequency f and terrain clearance angle 3. The method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion according to claim 1, characterized in that: In step S1, after collecting the measurement data and before dividing the data set, the measured values ​​of the confirmation hyperparameters and the sub-model prediction values ​​are scaled by normalization so that the measured values ​​of the confirmation hyperparameters and the sub-model prediction values ​​have the same contribution or effect in confirmation.

4. The method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion according to claim 1, characterized in that: The confirmation hyper parameters include signal frequency f, transceiver distance d, transmitting antenna height h1, receiving antenna height h2 and terrain clearance angle 5. The method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion according to claim 1, characterized in that: In step S1, the collected measurement data is divided into a training set, a validation set and a test set in a ratio of 6:2:

2.

6. The method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion according to claim 1, characterized in that: In step S2, the ITU-R P.1546 and 2001 models are used as sub-models, and a combined prediction model is established based on the entropy weight theory, including: Step 21. Construct the discriminant matrix: The values ​​in the discriminant matrix are composed of the absolute values ​​of the difference between the measured values ​​and the predicted values ​​of the sub-models; assuming that each frequency band has n measured values, the absolute value of the difference between the measured value and the predicted value at the j-th training point of the i-th independent prediction model is Lr ij ; Then the judgment matrix V is: Where, i = {1, 2}, i = 1 corresponds to the ITU-R P.1546 model, i = 2 corresponds to the ITU-R P.2001 model; j = {1, 2, ..., n}, corresponding to n measured values ​​at each frequency; Step 22. Calculate the weight p of the sub-model at each training point based on the value in the discriminant matrix ij ; Step 23. Based on the weight p of the sub-model at each training point ij Calculate the sub-model information entropy H i ; Step 24. Based on the sub-model information entropy H i Calculate the weight W of each sub-model in the combined prediction model i ; Step 25. Based on the weight W of each sub-model in the combined prediction model i , establish a combined prediction model, and use the combined prediction model to predict the path loss L C =w1×L1+w2×L2; Where, L c is the path loss, L1 is the value predicted by the ITU-R P.1546 model; W1 is the weight of the ITU-R P.1546 model; L2 is the value predicted by the ITU-R P.2001 model; W2 is the weight of the ITU-R P.2001 model.

7. The method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion according to claim 6, characterized in that: The absolute value of the difference between the measured value and the predicted value at the jth training point of the i-th independent prediction model is Lr ij Obtained by the following formula: Lr ij =|Lr obs -Lr pre |; Where Lr obs is the measured value, Lr pre is the predicted value.

8. The method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion according to claim 6, characterized in that: The sum of the weights of the two sub-models is equal to 1.

9. The method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion according to claim 1, characterized in that: The combined prediction model input is the signal frequency f, the longitude and latitude position of the transmitting antenna (L o1 ,L a1 ), the longitude and latitude position of the receiving antenna (L o2 ,L a2 ), ITU-R P.1546 and 2001 model predicted loss values, and the output is the final path loss L calculated by the combined prediction model c .

10. The method for constructing a VHF / UHF radio wave propagation prediction model based on entropy weight fusion according to claim 1, characterized in that: The combined prediction model is tested, adjusted and verified using the test set and validation set data so that the accuracy of the combined prediction model reaches the expected level, including: Use the combined prediction model to predict the test set data, and compare and analyze the predicted value with the measured value and the sub-model predicted value to determine whether the prediction accuracy of the combined prediction model meets expectations; if the prediction accuracy of the combined prediction model is lower than that of the prediction sub-model and does not meet expectations, replace the confirmation hyperparameters and repeat step S2; When the prediction accuracy of the combined prediction model reaches the expected level, the final prediction model is output, which is the VHF / UHF radio wave propagation combined prediction model based on entropy weight theory.

Images