Big Data-Based Train Travel Time Prediction Method and System

Abstract

The present invention relates to the technical field of intelligent transportation, and discloses a method and system for predicting the train passing time based on big data. Historical passing data records are obtained, a historical early passing time series and a historical late passing time series are determined, an early passing time smoothness coefficient, an early passing time jump wave coefficient, a late passing time smoothness coefficient, and a late passing time jump wave coefficient are calculated, and a historical passing time change factor is calculated. Historical driving speeds are collected, and a historical driving speed change factor is calculated. A train passing time prediction factor is determined according to the historical passing time change factor and the historical driving speed change factor. The real-time expected passing time is predicted according to the train passing time prediction factor to obtain the train passing prediction time, and the historical passing data records are analyzed intelligently, so as to improve the prediction accuracy and reliability, eliminate prediction errors, effectively avoid conflicts between vehicles and trains, and greatly improve the passing safety and efficiency of the port area roads.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent transportation, and more particularly, to a method and system for predicting train passing time based on big data. Background Art

[0002] The transportation methods of goods in the port area are rich and diverse, and train transportation is an extremely important part of them. As an important goods distribution center, the port area undertakes a large number of goods handling tasks. Trains play a key role in goods transportation with their advantages of large transportation volume and long-distance transportation. Their efficient operation has a decisive impact on the overall operation efficiency of the port area.

[0003] In the prior art, the prediction of train passing time relies on a conventional traffic prediction model constructed based on limited traffic flow data and simple road information. In the port area, there are not only frequent train passages, but also a large number of trucks and loading and unloading equipment shuttling in a complex road network. The road layout, train operation routes, goods loading and unloading operations, etc. are very different from the conventional scenarios, making it difficult to adapt to the complex and unique environment of the port area. Moreover, the conventional traffic prediction model ignores the rich information contained in the historical passing data of trains. Historical passing data can reflect some long-term laws and potential trends in the process of train passing, which has an important value for predicting the train passing time more comprehensively and accurately. These special circumstances make the existing methods unable to accurately predict the train passing time, and it is even more difficult to reasonably plan the traffic in the port area according to the train passing time. Summary of the Invention

[0004] Embodiments of the present invention provide a method and system for predicting train passing time based on big data. The present invention can perform intelligent analysis on the historical passing data and real-time passing data of trains, improve the prediction accuracy and reliability of train passing time, eliminate prediction errors, effectively avoid conflicts between vehicles and trains, and greatly improve the traffic safety and efficiency of the port area roads.

[0005] To achieve the above object, the present invention provides a method for predicting train passing time based on big data, including:

[0006] Determine the train to be predicted, obtain multiple historical passing data records of the train to be predicted, analyze each historical passing data record, and determine the historical early passing time sequence and the historical late passing time sequence;

[0007] Perform smoothing calculation and wave jump calculation on the historical early passing time sequence to obtain the early passing time smoothing coefficient and the early passing time wave jump coefficient corresponding to the historical early passing time sequence, and perform smoothing calculation and wave jump calculation on the historical late passing time sequence to obtain the late passing time smoothing coefficient and the late passing time wave jump coefficient corresponding to the historical late passing time sequence;

[0008] Calculate the historical passing time change factor of the train to be predicted based on the early passing time smoothing coefficient, the early passing time jump wave coefficient, the delayed passing time smoothing coefficient, and the delayed passing time jump wave coefficient;

[0009] Collect the historical driving speed corresponding to each historical passing data record, and calculate the historical driving speed change factor of the train to be predicted based on all the historical driving speeds;

[0010] Determine the train passing time prediction factor of the train to be predicted according to the historical passing time change factor and the historical driving speed change factor;

[0011] Pre-deploy information collection devices to collect the real-time driving information of the train to be predicted, determine the real-time expected passing time based on the real-time driving information, and perform prediction adjustment on the real-time expected passing time according to the train passing time prediction factor to obtain the train passing prediction time of the train to be predicted.

[0012] Further, when obtaining multiple historical passing data records of the train to be predicted and analyzing each historical passing data record to determine the historical early passing time sequence and the historical delayed passing time sequence, it includes:

[0013] Obtain the historical expected passing time corresponding to each historical passing data record;

[0014] Analyze each historical passing data record to determine the corresponding historical actual passing time, and generate a passing early mark, a passing expected mark, and a passing delayed mark for the historical passing data record according to the historical actual passing time and the historical expected passing time;

[0015] When the historical actual passing time is less than the historical expected passing time, generate the passing early mark for the corresponding historical passing data record;

[0016] When the historical actual passing time is equal to the historical expected passing time, generate the passing expected mark for the corresponding historical passing data record;

[0017] When the historical actual passing time is greater than the historical expected passing time, generate the passing delayed mark for the corresponding historical passing data record;

[0018] Extract the historical actual passing time corresponding to each passing early mark, calculate the early passing time difference between the historical actual passing time and the historical expected passing time, and determine the historical early passing time sequence according to all the early passing time differences;

[0019] Extract the historical actual passing time corresponding to each passing delay tag, calculate the delay passing time difference between the historical actual passing time and the historical expected passing time, and determine the historical delay passing time sequence according to all the delay passing time differences.

[0020] Further, when performing smoothing calculation and jump wave calculation on the historical early passing time sequence to obtain the early passing time smoothing coefficient and the early passing time jump wave coefficient corresponding to the historical early passing time sequence, it includes:

[0021] Calculate the early passing time smoothing coefficient corresponding to the historical early passing time sequence according to the following formula:

[0022] ;

[0023] where s1 is the early passing time smoothing coefficient corresponding to the historical early passing time sequence, a1 is the number of early passing time differences in the historical early passing time sequence, f d+1 is the (d + 1)-th early passing time difference in the historical early passing time sequence, f d is the d-th early passing time difference in the historical early passing time sequence;

[0024] Calculate the early passing time jump wave coefficient corresponding to the historical early passing time sequence according to the following formula:

[0025] ;

[0026] where s2 is the early passing time jump wave coefficient corresponding to the historical early passing time sequence.

[0027] Further, when performing smoothing calculation and jump wave calculation on the historical delay passing time sequence to obtain the delay passing time smoothing coefficient and the delay passing time jump wave coefficient corresponding to the historical delay passing time sequence, it includes:

[0028] Calculate the delay passing time smoothing coefficient corresponding to the historical delay passing time sequence according to the following formula:

[0029] ;

[0030] where s3 is the delay passing time smoothing coefficient corresponding to the historical delay passing time sequence, a2 is the number of delay passing time differences in the historical delay passing time sequence, h g+1 is the (g + 1)-th delay passing time difference in the historical delay passing time sequence, h g is the g-th delay passing time difference in the historical delay passing time sequence;

[0031] Calculate the delay passing time jump coefficient corresponding to the historical delay passing time series according to the following formula:

[0032] ;

[0033] where s4 is the delay passing time jump coefficient corresponding to the historical delay passing time series.

[0034] Further, when calculating the historical passing time change factor of the train to be predicted based on the early passing time smooth coefficient, the early passing time jump coefficient, the delay passing time smooth coefficient, and the delay passing time jump coefficient, it includes:

[0035] Determine the number of early passing marks of the early passing mark, determine the number of expected passing marks of the expected passing mark, and determine the number of delay passing marks of the delay passing mark;

[0036] Calculate the historical passing time change factor of the train to be predicted according to the following formula:

[0037] ;

[0038] where k is the historical passing time change factor of the train to be predicted, w1 is the number of early passing marks, w2 is the number of expected passing marks, and w3 is the number of delay passing marks.

[0039] Further, when collecting the historical driving speed corresponding to each historical passing data record and calculating the historical driving speed change factor of the train to be predicted according to all the historical driving speeds, it includes:

[0040] Randomly distribute all the historical driving speeds onto the blank dot plot template and determine the position where the maximum historical driving speed is located;

[0041] Determine the initial historical driving speed and the end historical driving speed on the blank dot plot template, and count the first quantity of the historical driving speeds between the initial historical driving speed and the maximum historical driving speed;

[0042] Count the second quantity of the historical driving speeds between the end historical driving speed and the maximum historical driving speed;

[0043] Judge whether both the first quantity and the second quantity are greater than the preset quantity. If so, divide the historical driving speeds between the initial historical driving speed and the maximum historical driving speed into the first historical driving speed sequence, where the first historical driving speed sequence does not include the maximum historical driving speed;

[0044] Divide the historical driving speeds between the terminal historical driving speed and the maximum historical driving speed into a second historical driving speed sequence, where the second historical driving speed sequence does not include the maximum historical driving speed;

[0045] If not, re-randomly allocate all the historical driving speeds until both the obtained first quantity and second quantity are greater than the preset quantity;

[0046] Calculate the historical driving speed change factor of the train to be predicted according to the first historical driving speed sequence and the second historical driving speed sequence.

[0047] Further, when calculating the historical driving speed change factor of the train to be predicted according to the first historical driving speed sequence and the second historical driving speed sequence, it includes:

[0048] Calculate the historical driving speed change factor of the train to be predicted according to the following formula:

[0049] ;

[0050] where e is the historical driving speed change factor of the train to be predicted, r1 is the initial historical driving speed, r2 is the maximum historical driving speed, r3 is the terminal historical driving speed, t1 is the first calculation coefficient, t2 is the second calculation coefficient, t3 is the third calculation coefficient, t1 + t2 + t3 = 1, t1 > 0, t2 > 0, t3 > 0, y is the number of historical driving speeds, u1 is the variance of the historical driving speeds other than the initial historical driving speed in the first historical driving speed sequence, u2 is the variance of the historical driving speeds other than the terminal historical driving speed in the second historical driving speed sequence, and u is the variance of all historical driving speeds other than the maximum historical driving speed.

[0051] Further, when determining the train passing time prediction factor of the train to be predicted according to the historical passing time change factor and the historical driving speed change factor, it includes:

[0052] Configure a first calculation weight for the historical passing time change factor and calculate the first product value of the historical passing time change factor and the first calculation weight;

[0053] Configure a second calculation weight for the historical driving speed change factor and calculate the second product value of the historical driving speed change factor and the second calculation weight, where the first calculation weight is greater than the second calculation weight;

[0054] Determine the sum value of the first product value and the second product value and use it as the train passing time prediction factor of the train to be predicted.

[0055] Further, when predicting and adjusting the real-time expected passing time according to the train passing time prediction factor to obtain the train passing prediction time of the train to be predicted, it includes:

[0056] Preset a plurality of preset train passing time prediction factors;

[0057] Preset a plurality of preset passing time prediction adjustment coefficients;

[0058] According to the relationship between the train passing time prediction factor and the preset train passing time prediction factor, select the corresponding preset passing time prediction adjustment coefficient, wherein the train passing time prediction factor and the preset passing time prediction adjustment coefficient are in a direct proportional relationship;

[0059] Calculate the product value of the selected preset passing time prediction adjustment coefficient and the real-time expected passing time as the train passing prediction time of the train to be predicted.

[0060] To achieve the above object, the present invention also provides a train passing time prediction system based on big data, including:

[0061] A sequence determination module, configured to determine the train to be predicted, obtain a plurality of historical passing data records of the train to be predicted, analyze each historical passing data record, and determine a historical early passing time sequence and a historical late passing time sequence;

[0062] A coefficient calculation module, configured to perform a smoothing calculation and a jump wave calculation on the historical early passing time sequence to obtain an early passing time smoothing coefficient and an early passing time jump wave coefficient corresponding to the historical early passing time sequence, and perform a smoothing calculation and a jump wave calculation on the historical late passing time sequence to obtain a late passing time smoothing coefficient and a late passing time jump wave coefficient corresponding to the historical late passing time sequence;

[0063] A first calculation module, configured to calculate a historical passing time change factor of the train to be predicted based on the early passing time smoothing coefficient, the early passing time jump wave coefficient, the late passing time smoothing coefficient, and the late passing time jump wave coefficient;

[0064] A second calculation module, configured to collect the historical driving speed corresponding to each historical passing data record, and calculate a historical driving speed change factor of the train to be predicted according to all the historical driving speeds;

[0065] A factor determination module, configured to determine the train passing time prediction factor of the train to be predicted according to the historical passing time change factor and the historical driving speed change factor;

[0066] The passage prediction module is used to pre-deploy information collection devices to collect the real-time driving information of the train to be predicted, determine the real-time expected passage time based on the real-time driving information, and perform prediction adjustment on the real-time expected passage time according to the train passage time prediction factor to obtain the train passage prediction time of the train to be predicted.

[0067] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0068] The present invention discloses a method and system for predicting train passage time based on big data. Historical passage data records are obtained, historical early passage time series and historical late passage time series are determined, the early passage time smoothing coefficient, early passage time jump coefficient, late passage time smoothing coefficient, and late passage time jump coefficient are calculated, and the historical passage time change factor is calculated. Historical driving speeds are collected, and the historical driving speed change factor is calculated. The train passage time prediction factor is determined according to the historical passage time change factor and the historical driving speed change factor. The real-time expected passage time is predicted according to the train passage time prediction factor to obtain the train passage prediction time. The historical passage data records are analyzed intelligently, the prediction accuracy and reliability are improved, the prediction error is eliminated, the conflict between vehicles and trains is effectively avoided, and the passage safety and efficiency of the port area roads are greatly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0070] Figure 1 shows a schematic flow chart of the method for predicting train passage time based on big data in an embodiment of the present invention;

[0071] Figure 2 shows a schematic structural diagram of the system for predicting train passage time based on big data in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0072] The following combines the drawings and embodiments to further describe in detail the specific embodiments of the present invention. The following embodiments are used to illustrate the present invention but are not used to limit the scope of the present invention.

[0073] In the description of the present application, it should be understood that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on the present application.

[0074] The terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present application, unless otherwise specified, the meaning of "a plurality" is two or more.

[0075] In the description of the present application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "coupled" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may be a mechanical connection or an electrical connection; it may be directly connected or indirectly connected through an intermediate medium, and it may be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to specific circumstances.

[0076] The following is a description of the preferred embodiments of the present invention in conjunction with the drawings.

[0077] As Figure 1 shown, an embodiment of the present invention discloses a method for predicting train passing time based on big data, including:

[0078] S110: Determine the train to be predicted, obtain a plurality of historical passing data records of the train to be predicted, analyze each historical passing data record, and determine a historical early passing time series and a historical late passing time series;

[0079] In this embodiment, the historical passing data record refers to the historical passing behavior of the train to be predicted, including passing time, traveling speed, position change, stopping stations, etc.

[0080] In this embodiment, the preferred number of historical passing data records obtained is 20, and it can be adjusted according to the actual situation specifically.

[0081] In some embodiments of the present application, when obtaining a plurality of historical passing data records of the train to be predicted, analyzing each historical passing data record, and determining a historical early passing time series and a historical late passing time series, it includes:

[0082] Obtain the historical expected passing time corresponding to each historical passing data record;

[0083] Analyze each historical passing data record to determine the corresponding historical actual passing time. Based on the historical actual passing time and the historical expected passing time, generate a passing-early mark, a passing-expected mark, and a passing-late mark for the historical passing data record;

[0084] When the historical actual passing time is less than the historical expected passing time, generate the passing-early mark for the corresponding historical passing data record;

[0085] When the historical actual passing time is equal to the historical expected passing time, generate the passing-expected mark for the corresponding historical passing data record;

[0086] When the historical actual passing time is greater than the historical expected passing time, generate the passing-late mark for the corresponding historical passing data record;

[0087] Extract the historical actual passing time corresponding to each passing-early mark, calculate the early passing time difference between the historical actual passing time and the historical expected passing time, and determine the historical early passing time sequence based on all the early passing time differences;

[0088] Extract the historical actual passing time corresponding to each passing-late mark, calculate the late passing time difference between the historical actual passing time and the historical expected passing time, and determine the historical late passing time sequence based on all the late passing time differences.

[0089] In this embodiment, the historical expected passing time is a time obtained based on the train running information when the train is passing, which belongs to a theoretical value. For example, the historical expected passing time is 15 minutes. The historical actual passing time refers to the time used by the train from the start of passing to the end of passing. The historical actual passing time and the historical expected passing time may be equal or not equal. For example, the historical actual passing time is 12 minutes.

[0090] In this embodiment, calculate the difference between the historical actual passing time and the historical expected passing time as the early passing time difference or the late passing time difference.

[0091] The beneficial effects of the above technical solution are as follows: The present invention determines the historical early passing time sequence based on all the early passing time differences and determines the historical late passing time sequence based on all the late passing time differences, realizing the fine division of the historical actual passing time and providing a prerequisite for the calculation of the historical passing time change factor.

[0092] S120: Perform smoothing calculation and spike calculation on the historical early passing time series to obtain the early passing time smoothing coefficient and the early passing time spike coefficient corresponding to the historical early passing time series. Perform smoothing calculation and spike calculation on the historical late passing time series to obtain the late passing time smoothing coefficient and the late passing time spike coefficient corresponding to the historical late passing time series;

[0093] In some embodiments of the present application, when performing smoothing calculation and spike calculation on the historical early passing time series to obtain the early passing time smoothing coefficient and the early passing time spike coefficient corresponding to the historical early passing time series, it includes:

[0094] Calculate the early passing time smoothing coefficient corresponding to the historical early passing time series according to the following formula:

[0095] ;

[0096] where s1 is the early passing time smoothing coefficient corresponding to the historical early passing time series, a1 is the number of early passing time differences in the historical early passing time series, f d+1 is the (d + 1)-th early passing time difference in the historical early passing time series, and f d is the d-th early passing time difference in the historical early passing time series;

[0097] Calculate the early passing time spike coefficient corresponding to the historical early passing time series according to the following formula:

[0098] ;

[0099] where s2 is the early passing time spike coefficient corresponding to the historical early passing time series.

[0100] The beneficial effects of the above technical solutions are as follows: The present invention obtains the early passing time smoothing coefficient and the early passing time spike coefficient according to the historical early passing time series, ensuring the calculation accuracy of the early passing time smoothing coefficient and the early passing time spike coefficient. Through two calculation methods, the comprehensiveness of the calculation is ensured. It can not only perform smoothing calculation on the historical early passing time series, that is, sequential calculation, to reflect the law of historical actual passing time, but also perform spike calculation on the historical early passing time series, that is, holistic calculation, to further reflect the law of historical actual passing time.

[0101] In some embodiments of the present application, when performing smoothing calculation and spike calculation on the historical late passing time series to obtain the late passing time smoothing coefficient and the late passing time spike coefficient corresponding to the historical late passing time series, it includes:

[0102] Calculate the delay passage time smoothing coefficient corresponding to the historical delay passage time series according to the following formula:

[0103] ;

[0104] where s3 is the delay passage time smoothing coefficient corresponding to the historical delay passage time series, a2 is the number of delay passage time differences in the historical delay passage time series, and h g+1 is the (g + 1)-th delay passage time difference in the historical delay passage time series, and h g is the g-th delay passage time difference in the historical delay passage time series;

[0105] Calculate the delay passage time jump coefficient corresponding to the historical delay passage time series according to the following formula:

[0106] ;

[0107] where s4 is the delay passage time jump coefficient corresponding to the historical delay passage time series.

[0108] S130: Calculate the historical passage time change factor of the train to be predicted based on the early passage time smoothing coefficient, the early passage time jump coefficient, the delay passage time smoothing coefficient, and the delay passage time jump coefficient;

[0109] In some embodiments of the present application, when calculating the historical passage time change factor of the train to be predicted based on the early passage time smoothing coefficient, the early passage time jump coefficient, the delay passage time smoothing coefficient, and the delay passage time jump coefficient, it includes:

[0110] Determine the number of early passage markers of the early passage marker, determine the number of expected passage markers of the expected passage marker, and determine the number of delay passage markers of the delay passage marker;

[0111] Calculate the historical passage time change factor of the train to be predicted according to the following formula:

[0112] ;

[0113] where k is the historical passage time change factor of the train to be predicted, w1 is the number of early passage markers, w2 is the number of expected passage markers, and w3 is the number of delay passage markers.

[0114] The beneficial effects of the above technical solution are as follows: According to the early passing time smoothing coefficient, the early passing time jump coefficient, the delayed passing time smoothing coefficient, and the delayed passing time jump coefficient, the present invention calculates the historical passing time change factor of the train to be predicted, ensuring the calculation accuracy and efficiency of the historical passing time change factor. The historical passing time change factor comprehensively reflects the change law and trend of the historical passing time. At the same time, by the number of early passing marks, the number of expected passing marks, and the number of delayed passing marks, the constrained calculation of the historical passing time change factor is realized, further ensuring the analysis accuracy and eliminating errors.

[0115] S140: Collect the historical driving speed corresponding to each historical passing data record, and calculate the historical driving speed change factor of the train to be predicted according to all the historical driving speeds;

[0116] In this embodiment, during the train's driving process, the driving speed changes. Therefore, the average value of the historical driving speeds corresponding to each historical passing data record is taken as the historical driving speed corresponding to each historical passing data record.

[0117] In some embodiments of the present application, when collecting the historical driving speed corresponding to each historical passing data record and calculating the historical driving speed change factor of the train to be predicted according to all the historical driving speeds, it includes:

[0118] Randomly allocate all the historical driving speeds to the blank dot map template, and determine the position where the maximum historical driving speed is located;

[0119] Determine the initial historical driving speed and the terminal historical driving speed on the blank dot map template, and count the first quantity of the historical driving speeds between the initial historical driving speed and the maximum historical driving speed;

[0120] Count the second quantity of the historical driving speeds between the terminal historical driving speed and the maximum historical driving speed;

[0121] Judge whether both the first quantity and the second quantity are greater than the preset quantity. If so, divide the historical driving speeds between the initial historical driving speed and the maximum historical driving speed into the first historical driving speed sequence, where the first historical driving speed sequence does not include the maximum historical driving speed;

[0122] Divide the historical driving speeds between the terminal historical driving speed and the maximum historical driving speed into the second historical driving speed sequence, where the second historical driving speed sequence does not include the maximum historical driving speed;

[0123] Otherwise, re - randomly allocate all historical driving speeds until both the obtained first quantity and second quantity are greater than the preset quantity;

[0124] Calculate the historical driving speed change factor of the train to be predicted according to the first historical driving speed sequence and the second historical driving speed sequence.

[0125] In this embodiment, the blank scatter plot template is a pre - designed scatter plot framework that does not contain any data series but has a basic chart structure and format settings. It has a horizontal axis (X - axis) and a vertical axis (Y - axis), with scales on the axes to determine the positions of data points. There are no specific numerical annotations on the vertical axis in the blank template, and the numerical values on the horizontal axis show an arithmetic progression, such as 1, 2, 3, 4, 5, etc., which is convenient for the corresponding annotation of historical driving speeds.

[0126] In this embodiment, by randomly allocating all historical driving speeds onto the blank scatter plot template, a scatter plot of historical driving speeds can be obtained.

[0127] In this embodiment, the initial historical driving speed refers to the first randomly allocated historical driving speed on the scatter plot, and the end - point historical driving speed refers to the last randomly allocated historical driving speed on the scatter plot.

[0128] In this embodiment, if the number of the maximum historical driving speeds is greater than or equal to 2, randomly select one of the maximum historical driving speeds and determine its position.

[0129] In this embodiment, the preset quantity is preferably 7, and can be adjusted according to the actual situation.

[0130] In this embodiment, when counting the first quantity, count the number of the initial historical driving speeds at the same time, but do not count the number of the maximum historical driving speeds. When counting the second quantity, count the number of the end - point historical driving speeds at the same time, but do not count the number of the maximum historical driving speeds.

[0131] The beneficial effects of the above - mentioned technical solution are as follows: The present invention calculates the historical driving speed change factor of the train to be predicted according to the first historical driving speed sequence and the second historical driving speed sequence, which ensures the calculation accuracy of the historical driving speed change factor, eliminates the need for manual calculation, eliminates the subjective error in calculation, and at the same time, the historical driving speed change factor can reflect the historical driving speed law of the train to be predicted, providing a basis for calculating the train passing - time prediction factor.

[0132] In some embodiments of the present application, when calculating the historical driving speed change factor of the train to be predicted according to the first historical driving speed sequence and the second historical driving speed sequence, it includes:

[0133] Calculate the historical driving speed change factor of the train to be predicted according to the following formula:

[0134] ;

[0135] where e is the historical driving speed change factor of the train to be predicted, r1 is the initial historical driving speed, r2 is the maximum historical driving speed, r3 is the terminal historical driving speed, t1 is the first calculation coefficient, t2 is the second calculation coefficient, t3 is the third calculation coefficient, t1 + t2 + t3 = 1, t1 > 0, t2 > 0, t3 > 0, y is the number of historical driving speeds, u1 is the variance of the historical driving speeds other than the initial historical driving speed in the first historical driving speed sequence, u2 is the variance of the historical driving speeds other than the terminal historical driving speed in the second historical driving speed sequence, and u is the variance of all historical driving speeds other than the maximum historical driving speed.

[0136] S150: Determine the train passing time prediction factor of the train to be predicted according to the historical passing time change factor and the historical driving speed change factor;

[0137] In some embodiments of the present application, when determining the train passing time prediction factor of the train to be predicted according to the historical passing time change factor and the historical driving speed change factor, it includes:

[0138] Configure a first calculation weight for the historical passing time change factor and calculate a first product value of the historical passing time change factor and the first calculation weight;

[0139] Configure a second calculation weight for the historical driving speed change factor and calculate a second product value of the historical driving speed change factor and the second calculation weight, where the first calculation weight is greater than the second calculation weight;

[0140] Determine the sum value of the first product value and the second product value and use it as the train passing time prediction factor of the train to be predicted.

[0141] In this embodiment, the first calculation weight is preferably 0.7 and the second calculation weight is preferably 0.3, and can be specifically adjusted adaptively according to the actual situation.

[0142] The beneficial effects of the above technical solutions are as follows: The present invention determines the train passing time prediction factor of the train to be predicted according to the historical passing time change factor and the historical driving speed change factor, realizes the comprehensive analysis of historical passing data records, and lays a foundation for the prediction of the train passing time of the train to be predicted.

[0143] S160: Pre-deploy information collection devices to collect the real-time driving information of the train to be predicted, determine the real-time expected passing time based on the real-time driving information, and perform prediction adjustment on the real-time expected passing time according to the train passing time prediction factor to obtain the train passing prediction time of the train to be predicted.

[0144] In this embodiment, the information collection devices include cameras, geomagnetic sensors, RFID devices, etc.

[0145] In this embodiment, the real-time driving information includes train images, driving speed, position, identity information, etc.

[0146] In some embodiments of the present application, when performing prediction adjustment on the real-time expected passing time according to the train passing time prediction factor to obtain the train passing prediction time of the train to be predicted, it includes:

[0147] Preset multiple preset train passing time prediction factors in advance;

[0148] Preset multiple preset passing time prediction adjustment coefficients in advance;

[0149] According to the relationship between the train passing time prediction factor and the preset train passing time prediction factor, select the corresponding preset passing time prediction adjustment coefficient, where the train passing time prediction factor and the preset passing time prediction adjustment coefficient are in a direct proportional relationship;

[0150] Calculate the product value of the selected preset passing time prediction adjustment coefficient and the real-time expected passing time as the train passing prediction time of the train to be predicted.

[0151] In this embodiment, it is preferably to preset three preset train passing time prediction factors in advance, including the first preset train passing time prediction factor, the second preset train passing time prediction factor, and the third preset train passing time prediction factor, and the first preset train passing time prediction factor is preferably 5, the second preset train passing time prediction factor is preferably 8, and the third preset train passing time prediction factor is preferably 11. Specifically, it can also be adjusted according to the actual situation.

[0152] In this embodiment, it is preferably to preset four preset passing time prediction adjustment coefficients in advance, including the first preset passing time prediction adjustment coefficient, the second preset passing time prediction adjustment coefficient, the third preset passing time prediction adjustment coefficient, and the fourth preset passing time prediction adjustment coefficient, and the first preset passing time prediction adjustment coefficient is preferably 0.85, the second preset passing time prediction adjustment coefficient is preferably 0.95, the third preset passing time prediction adjustment coefficient is preferably 1.05, and the fourth preset passing time prediction adjustment coefficient is preferably 1.15. Specifically, it can also be adjusted according to the actual situation.

[0153] In this embodiment, when the train passing time prediction factor is less than the first preset train passing time prediction factor, the product value of the first preset passing time prediction adjustment coefficient and the real-time expected passing time is calculated as the train passing prediction time of the train to be predicted. For example, if the real-time expected passing time is 16 minutes, the train passing prediction time is 13.6 minutes.

[0154] In this embodiment, when the train passing time prediction factor is greater than or equal to the first preset train passing time prediction factor and less than the second preset train passing time prediction factor, the product value of the second preset passing time prediction adjustment coefficient and the real-time expected passing time is calculated as the train passing prediction time of the train to be predicted.

[0155] In this embodiment, when the train passing time prediction factor is greater than or equal to the second preset train passing time prediction factor and less than the third preset train passing time prediction factor, the product value of the third preset passing time prediction adjustment coefficient and the real-time expected passing time is calculated as the train passing prediction time of the train to be predicted.

[0156] In this embodiment, when the train passing time prediction factor is greater than or equal to the third preset train passing time prediction factor, the product value of the fourth preset passing time prediction adjustment coefficient and the real-time expected passing time is calculated as the train passing prediction time of the train to be predicted.

[0157] The beneficial effects of the above technical solutions are as follows: The present invention realizes the accurate determination of the train passing prediction time, effectively avoids the conflict between vehicles and trains, greatly improves the traffic safety and efficiency of the port area roads, provides intuitive and accurate passing prompts for drivers, accelerates the traffic fluency of the port area, and improves the overall production efficiency of the wharf.

[0158] In order to further elaborate the technical idea of the present invention, the technical solutions of the present invention will be described in combination with specific application scenarios.

[0159] Correspondingly, as Figure 2 shown, the present application also provides a big data-based train passing time prediction system, including:

[0160] A sequence determination module, configured to determine the train to be predicted, obtain multiple historical passing data records of the train to be predicted, analyze each historical passing data record, and determine a historical early passing time sequence and a historical late passing time sequence;

[0161] A coefficient calculation module is configured to perform smoothing calculation and wave-jump calculation on the historical early passing time series to obtain an early passing time smoothing coefficient and an early passing time wave-jump coefficient corresponding to the historical early passing time series, and perform smoothing calculation and wave-jump calculation on the historical late passing time series to obtain a late passing time smoothing coefficient and a late passing time wave-jump coefficient corresponding to the historical late passing time series;

[0162] A first calculation module is configured to calculate a historical passing time change factor of the train to be predicted based on the early passing time smoothing coefficient, the early passing time wave-jump coefficient, the late passing time smoothing coefficient, and the late passing time wave-jump coefficient;

[0163] A second calculation module is configured to collect the historical driving speed corresponding to each historical passing data record, and calculate a historical driving speed change factor of the train to be predicted according to all the historical driving speeds;

[0164] A factor determination module is configured to determine a train passing time prediction factor of the train to be predicted according to the historical passing time change factor and the historical driving speed change factor;

[0165] A passing prediction module is configured to pre-deploy information collection devices, collect real-time driving information of the train to be predicted, determine a real-time expected passing time based on the real-time driving information, and perform prediction adjustment on the real-time expected passing time according to the train passing time prediction factor to obtain a train passing prediction time of the train to be predicted.

[0166] In the description of the above embodiments, specific features, structures, materials, or characteristics may be combined in any one or more embodiments or examples in a suitable manner.

[0167] Although the present invention has been described above with reference to embodiments, various improvements can be made to it and components therein can be replaced with equivalents without departing from the scope of the present invention. In particular, as long as there is no structural conflict, the various features in the embodiments disclosed in the present invention can be combined with each other in any way, and only for the sake of saving space and resources, the description of all these combinations is not given in this specification.

[0168] Those of ordinary skill in the art can understand that the above are only preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A train travel time prediction method based on big data, characterized in that: include: Determine a train to be predicted, obtain multiple historical travel data records of the train to be predicted, analyze each historical travel data record, and determine a historical advance travel time sequence and a historical delay travel time sequence; Performing smoothing calculation and wave skipping calculation on the historical advance passage time sequence to obtain the advance passage time smoothing coefficient and the advance passage time wave skipping coefficient corresponding to the historical advance passage time sequence, and performing smoothing calculation and wave skipping calculation on the historical delayed passage time sequence to obtain the delayed passage time smoothing coefficient and the delayed passage time wave skipping coefficient corresponding to the historical delayed passage time sequence; Calculate the historical travel time variation factor of the train to be predicted based on the advance travel time smoothing coefficient, the advance travel time skipping coefficient, the delayed travel time smoothing coefficient and the delayed travel time skipping coefficient; Collecting the historical travel speed corresponding to each historical travel data record, and calculating the historical travel speed change factor of the train to be predicted based on all the historical travel speeds; Determine a train travel time prediction factor of the train to be predicted according to the historical travel time change factor and the historical travel speed change factor; Predeploy information collection equipment to collect real-time travel information of the train to be predicted, determine the real-time expected travel time based on the real-time travel information, and predict and adjust the real-time expected travel time according to the train travel time prediction factor to obtain the train travel prediction time of the train to be predicted.

2. The train travel time prediction method based on big data according to claim 1 is characterized in that: When obtaining a plurality of historical travel data records of the train to be predicted, analyzing each historical travel data record, and determining a historical advance travel time sequence and a historical delay travel time sequence, the method includes: Obtain the historical expected travel time corresponding to each historical travel data record; Analyze each historical travel data record to determine the corresponding historical actual travel time, and generate a travel advance mark, a travel expectation mark, and a travel delay mark for the historical travel data record according to the historical actual travel time and the historical expected travel time; When the historical actual travel time is less than the historical expected travel time, the advance travel mark is generated for the corresponding historical travel data record; When the historical actual travel time is equal to the historical expected travel time, the travel expectation mark is generated for the corresponding historical travel data record; When the historical actual travel time is greater than the historical expected travel time, the travel delay mark is generated for the corresponding historical travel data record; Extracting the historical actual travel time corresponding to each advance travel mark, calculating the advance travel time difference between the historical actual travel time and the historical expected travel time, and determining the historical advance travel time sequence according to all the advance travel time differences; The historical actual travel time corresponding to each delayed travel mark is extracted, the delayed travel time difference between the historical actual travel time and the historical expected travel time is calculated, and the historical delayed travel time sequence is determined according to all the delayed travel time differences.

3. The train travel time prediction method based on big data according to claim 2 is characterized in that: When performing smoothing calculation and wave skipping calculation on the historical advance passage time sequence to obtain the advance passage time smoothing coefficient and the advance passage time wave skipping coefficient corresponding to the historical advance passage time sequence, it includes: The advance travel time smoothing coefficient corresponding to the historical advance travel time series is calculated according to the following formula: ； Among them, s1 is the advance travel time smoothing coefficient corresponding to the historical advance travel time series, a1 is the number of advance travel time differences in the historical advance travel time series, and f d+1 is the d+1th advance travel time difference in the historical advance travel time series, f d is the dth advance passage time difference in the historical advance passage time series; The advance travel time jump coefficient corresponding to the historical advance travel time series is calculated according to the following formula: ； Among them, s2 is the advance passage time jump coefficient corresponding to the historical advance passage time series.

4. The train travel time prediction method based on big data according to claim 3 is characterized in that: When performing smoothing calculation and wave skipping calculation on the historical delayed passage time series to obtain a delayed passage time smoothing coefficient and a delayed passage time wave skipping coefficient corresponding to the historical delayed passage time series, it includes: The delayed travel time smoothing coefficient corresponding to the historical delayed travel time series is calculated according to the following formula: ； Among them, s3 is the delayed travel time smoothing coefficient corresponding to the historical delayed travel time series, a2 is the number of delayed travel time differences in the historical delayed travel time series, and h g+1 is the g+1th delayed travel time difference in the historical delayed travel time series, h g is the g-th delayed passage time difference in the historical delayed passage time series; The delayed travel time jump coefficient corresponding to the historical delayed travel time series is calculated according to the following formula: ； Among them, s4 is the delayed passage time jump coefficient corresponding to the historical delayed passage time series.

5. The method for predicting train travel time based on big data according to claim 4, characterized in that: When calculating the historical travel time variation factor of the train to be predicted based on the advance travel time smoothing coefficient, the advance travel time skipping coefficient, the delayed travel time smoothing coefficient and the delayed travel time skipping coefficient, it includes: Determine the number of advance pass marks of the advance pass mark, determine the number of expected pass marks of the expected pass mark, and determine the number of delayed pass marks of the delayed pass mark; The historical travel time variation factor of the train to be predicted is calculated according to the following formula: ； Among them, k is the historical travel time change factor of the train to be predicted, w1 is the number of advance travel marks, w2 is the number of expected travel marks, and w3 is the number of delayed travel marks.

6. The train travel time prediction method based on big data according to claim 1 is characterized in that: When collecting the historical travel speed corresponding to each historical travel data record and calculating the historical travel speed change factor of the train to be predicted according to all the historical travel speeds, it includes: All historical driving speeds are randomly assigned to the blank point map template, and the location of the maximum historical driving speed is determined; Determine an initial historical driving speed and a terminal historical driving speed on the blank point map template, and count a first number of historical driving speeds between the initial historical driving speed and the maximum historical driving speed; Counting a second number of historical driving speeds between the terminal historical driving speed and the maximum historical driving speed; determining whether both the first number and the second number are greater than a preset number, and if so, dividing the historical driving speeds between the initial historical driving speed and the maximum historical driving speed into a first historical driving speed sequence, wherein the first historical driving speed sequence does not include the maximum historical driving speed; dividing the historical driving speeds between the terminal historical driving speed and the maximum historical driving speed into a second historical driving speed sequence, wherein the second historical driving speed sequence does not include the maximum historical driving speed; If not, all historical driving speeds are randomly redistributed until both the first number and the second number obtained are greater than the preset number; The historical running speed change factor of the train to be predicted is calculated according to the first historical running speed sequence and the second historical running speed sequence.

7. The train travel time prediction method based on big data according to claim 6 is characterized in that: When calculating the historical travel speed change factor of the train to be predicted according to the first historical travel speed sequence and the second historical travel speed sequence, it includes: The historical speed change factor of the train to be predicted is calculated according to the following formula: ； Wherein, e is the historical speed change factor of the train to be predicted, r1 is the initial historical speed, r2 is the maximum historical speed, r3 is the terminal historical speed, t1 is the first calculation coefficient, t2 is the second calculation coefficient, t3 is the third calculation coefficient, t1+t2+t3=1, t1>0, t2>0, t3>0, y is the number of historical speeds, u1 is the variance of the historical speeds other than the initial historical speed in the first historical speed sequence, u2 is the variance of the historical speeds other than the terminal historical speed in the second historical speed sequence, and u is the variance of all historical speeds other than the maximum historical speed.

8. The train travel time prediction method based on big data according to claim 1 is characterized in that: When determining the train transit time prediction factor of the train to be predicted according to the historical transit time variation factor and the historical travel speed variation factor, it includes: configuring a first calculation weight for the historical travel time change factor, and calculating a first product value of the historical travel time change factor and the first calculation weight; configuring a second calculation weight for the historical driving speed change factor, and calculating a second product value of the historical driving speed change factor and the second calculation weight, wherein the first calculation weight is greater than the second calculation weight; A sum of the first product value and the second product value is determined and used as a train travel time prediction factor of the train to be predicted.

9. The train travel time prediction method based on big data according to claim 1, characterized in that: When the real-time expected travel time is predicted and adjusted according to the train travel time prediction factor to obtain the train travel prediction time of the train to be predicted, it includes: Presetting a plurality of preset train travel time prediction factors; Presetting a plurality of preset travel time prediction adjustment coefficients; According to the relationship between the train travel time prediction factor and the preset train travel time prediction factor, a corresponding preset travel time prediction adjustment coefficient is selected, wherein the train travel time prediction factor and the preset travel time prediction adjustment coefficient are in a positive proportional relationship; The product value of the selected preset travel time prediction adjustment coefficient and the real-time expected travel time is calculated as the train travel prediction time of the train to be predicted.

10. A train travel time prediction system based on big data, applied to the train travel time prediction method based on big data as claimed in any one of claims 1 to 9, characterized in that: include: A sequence determination module is used to determine a train to be predicted, obtain multiple historical travel data records of the train to be predicted, analyze each historical travel data record, and determine a historical advance travel time sequence and a historical delay travel time sequence; A coefficient calculation module is used to perform smoothing calculation and wave skipping calculation on the historical advance passage time sequence to obtain the advance passage time smoothing coefficient and the advance passage time wave skipping coefficient corresponding to the historical advance passage time sequence, and perform smoothing calculation and wave skipping calculation on the historical delayed passage time sequence to obtain the delayed passage time smoothing coefficient and the delayed passage time wave skipping coefficient corresponding to the historical delayed passage time sequence; A first calculation module is used to calculate the historical travel time variation factor of the train to be predicted based on the advance travel time smoothing coefficient, the advance travel time skipping coefficient, the delayed travel time smoothing coefficient and the delayed travel time skipping coefficient; The second calculation module is used to collect the historical travel speed corresponding to each historical travel data record, and calculate the historical travel speed change factor of the train to be predicted according to all the historical travel speeds; A factor determination module, used to determine the train travel time prediction factor of the train to be predicted according to the historical travel time change factor and the historical travel speed change factor; The travel prediction module is used to pre-deploy information collection equipment to collect real-time travel information of the train to be predicted, determine the real-time expected travel time based on the real-time travel information, and predict and adjust the real-time expected travel time according to the train travel time prediction factor to obtain the train travel prediction time of the train to be predicted.

Images