MXPA00006687A - System and method for classifying and tracking aircraft and vehicles on the grounds of an airport - Google Patents

Abstract

A system and method for classifying and tracking aircraft and ground vehicles on the grounds of an airport is provided. The system generates a signal as a vehicle passes over a sensor under the taxiways or runways and analyzes the signal to determine a signature for the vehicle and the estimated speed of the vehicle. The determined signature and estimated speed may be used to track the movement of the vehicle around the grounds of the airport.

Description

SYSTEM AND METHOD FOR CLASSIFYING AND TRACKING AIRCRAFT AND VEHICLES ON THE GROUNDS OF A AIRPORT DESCRIPTION OF THE INVENTION This invention relates in general to a system and method for tracking vehicles, such as aircraft or land transportation vehicles, in the field and in particular to a system and method for classifying and tracking aircraft or other vehicles, while they are in the facilities of an airport. It is desirable to be able to track aircraft, both in the air and on the ground in order to avoid collisions between aircraft either in the air or on the ground. It is well known that aircraft are tracked as long as they are in the air by air traffic controllers using the radar. These systems effectively track the aircraft while it is in the air and ensure that the aircraft does not collide during the flight. However, these radar systems can not be easily used to track the aircraft on land, because the radar is typically mounted on the ground and radiates its energy upwards. Therefore, another extreme is necessary to track aircraft while on the ground. It is also desirable to have the ability to track the aircraft or other vehicles while the aircraft and vehicles are in the airport facilities as many collisions and accidents occur while the aircraft is rolling over the runway, landing or taking off. In particular, an aircraft is more vulnerable while on the ground, since radar can not help avoid collisions with other ships and there are also other ground transportation vehicles that can collide with the aircraft. A conventional system for guiding the aircraft already equipped with Instrument Landing Systems (ILS) along the circulation tracks uses the inductive circuits installed in the right half and the left half of the ci culation track. Each inductive circuit is driven at a different frequency so that a sensor mounted inside the aircraft detects the combined magnetic field induced by the inductive circuits and supplies the resulting ILS signal to provide the right / left guidance instructions for the pilot of the aircraft. through the ILS system. In addition, a dipole initiating antenna is embedded in the take-off runway or the circulation track and transmits a signal with a unique identifier so that the equipment (ATC) on the aircraft can detect the unique identifier and send information to an installation of air traffic controller which can determine the position of the aircraft. This conventional system is expensive because it requires an inductive circuit to guide the aircraft as well as a dipole antenna to determine the position of the aircraft. In addition, this conventional system can not automatically track an aircraft without communication between the aircraft and air traffic controllers (ATC) and can not automatically determine the type of aircraft. This conventional system also requires a typical ILS system on board the aircraft in order to track and guide the aircraft. This is especially a problem for smaller aircraft that can not have the ILS equipment installed as it is not required for smaller aircraft. This system also requires that some equipment be added to the aircraft, which means that only the aircraft that has been modified properly can be used with the system. It is also known that automobiles can be classified using an inductive circuit system. This system can also be capable of determining the type of car passing on the inductive circuit as well as the determined speed of the car. However, this system can not be used to track and classify aircraft as well as other vehicles, since a different technique to classify aircraft as opposed to automobiles is necessary and this conventional system does not really intend to track cars as cars move along the path . Therefore, there is a need for the system and method for classifying and tracking aircraft or other vehicles on the ground of an airport that avoids these or other problems of known systems and methods., and it is to this end that the present invention is focused. The invention provides a system for classifying and tracking aircraft and land vehicles in the facilities of an airport that can provide an air traffic controller with knowledge of where each aircraft or vehicle is on the runway surfaces or the runway. in any weather conditions. In addition, the system can combine other existing sensors, such as the airborne monitoring radar within the system to provide an integrated system with air and ground coverage. The system may also alert an air traffic controller (ATC) when a potential collision may occur between objects on the ground. The system, due to the classification of the aircraft, can provide an exit runway for the ATC with security intervals between aircraft. Using the existing data about each aircraft, such as the airline and the flight number, the aircraft that is tracked can have legends associated with it, indicating the flight line and flight number. The system with the inductive circuit sensors is less expensive than a conventional system. The system can also be integrated with a radar to provide an air and ground tracking system. The system can also be extended as the size of the airport expands, due to the addition of additional sensors due to taxiways and traffic. The system operates in all weather and visibility conditions and is very reliable since the system automatically checks its operation, reports any faults to an operator, and there are no moving parts in the system for wear. The system can also be installed quickly. For example, on new runways, the sensors are embedded directly in the paving material on the surface. The system can be quickly integrated into the airport's air traffic control system. Further. since the system has a plurality of inductive circuit sensors connected to the distributed electronic sensor units, there is no single point of failure as exists with conventional complex radar systems. The portion of the system located near the runways or bearing is sealed inside a waterproof enclosure that can be powered by solar cells or by direct energy (AC or DC). According to the invention, a system is provided for classifying and tracking vehicles in the facilities of an airport, the airport having at least one runway and at least one runway. The system comprises one or more sensors placed at predetermined locations adjacent to the take-off runway and the bearing track to generate a signal when a vehicle passes over a particular sensor, means for selecting one of a predetermined number of classifications for the passing vehicle. on the particular sensor based on the signal generated by the particular sensor, means for determining the position of the vehicle having the selected classification on the airport facilities as the vehicle passes over one or more sensors, and means for displaying a vehicle representation at the airport facilities, the representation of the vehicle indicating the selected classification of the vehicle and its placement in the airport facilities. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a diagram illustrating the layout of a common airport that is displayed on an aircraft tracking and classification system according to the invention: Figure 2 is a block diagram illustrating the system for the classification and tracking of aircraft according to the invention; Figure 3 is a block diagram illustrating more details of the system for the classification and tracking of aircraft shown in Figure 2; Figure 4 is a block diagram illustrating the details of the inductive circuit system shown in Figure 3; Figure 5 is a block diagram illustrating more details of the waveform detection and extraction portion of the system shown in Figure 4; Figure 6 is a diagram illustrating an example of the signal detection according to the invention; Figure 7 is a block diagram illustrating an example of a classifier, such as an elliptical base function network for classifying the aircraft according to the invention; Figure 8 is a block diagram illustrating an example of the activation function for the elliptical base function network type of classifier of Figure 7; Figure 9 is a block diagram illustrating more details of the central computer portion of the system of Figure 3; Figure 10 is a block diagram illustrating more details of the data fusion and the tracking portion of the central computer portion shown in Figure 9; and Figure 11 is a diagram illustrating an example of the tracking displayed according to the invention. The invention is particularly applicable to a system and method for classifying and tracking aircraft in the facilities of an airport. It is in this context that the invention will be described. However, it will be appreciated that the system and method according to the invention have greater utility. For example, the system can also track and classify land vehicles in the facilities of an airport. Figure 1 is a diagram illustrating an example of a simple distribution of a small airport 20 and an exit runway 22 that is displayed on the aircraft classification and tracking system according to the invention. The invention can be extended to handle any airport size. In the example shown, the airport has a main runway 24 and one or more runways 26, a control tower 28 and a terminal 30. The runway shown in this example has three aircraft that are tracked including a Boeing 757 (B757) 32 that is on the runway, a Boeing 747 (B747) that is on the runway and a Cesna 172 (C172) that is also on the runway. As the B757 32 takes off from the take-off runway, the B747 will move on the take-off runway and the runway will be automatically updated according to the invention as described below. Furthermore, according to the invention, the aircraft type such as the Boeing 747. Boeing 757 or Cessna 172, can be determined automatically by the system according to the invention without any communication with the aircraft as described below. In addition, the tracking and classification of the aircraft occurs without additional equipment on the aircraft so that the system can be used with any aircraft and does not require a specially modified aircraft. The aircraft sorting and tracking system according to the invention will now be described.

Figure 2 is a diagram illustrating an aircraft classification and tracking system 40 according to the invention. The system 40 may include one or more inductive circuit sensors 42 which is positioned at several points below the runways and runways of the airport. For new airports, inductive circuit sensors can be embedded directly in the paving material of runways and runways during the construction of the airport. One or more inductive circuit sensors are connected to an electronic sensor unit 44 that detects the aircraft and other vehicles passing over the inductive circuit sensors and also determines the type of aircraft and vehicles (i.e., classifies the aircraft and vehicles) that pass over the inductive circuit sensors as described below. The output of the sensor electronic component unit may be combined with a signal from an approach radius 46 and communicated to a central computer 48 which may be located within the air traffic control tower of the airport. The central computer 48 may include a communications system 50, a data fusion and tracking system 52 and a display unit 54. The output of the display unit, as shown in Figure 1 is also shown in this diagram. . The communication system 50 receives the signals from the electronic components of the sensor and the approach radar and processes them as described below. The data fusion and tracking system takes the information generated by one or more inductive circuit sensors and other inputs, such as an approach radar, and then generates the tracking information for aircraft and vehicles on the runways and lanes. bearing tracks of the airport. For example, based on the ratings made by more than one of the inductive circuit sensors, the data fusion and tracking system can determine the information from each of the inductive circuit sensors that refer to it. aircraft and then merge the different data into a single data record with respect to that aircraft. Therefore, the data fusion and tracking system associates the inductive circuit sensor data about the same aircraft. The data fusion and tracking system may also have raid logic to alert the user about a potential collision. The data fusion and tracking system will be described in greater detail with reference to Figures 9 and 10. The display unit presents the results of the data fusion and tracking system as shown in greater detail in Figure 1. The operation of the aircraft classification and tracking system will now be briefly described. As shown in Figure 2, as an aircraft, such as McDonell Douglas L1011 (MD11) 56 or a Cessna 172 58, passes over an inductive circuit sensor 42, the inductive circuit sensor generates a signal as described below. This signal, as shown in Figure 2, is almost unique for each type of aircraft since each type of aircraft has a different size and profile. These signals from the inductive circuit sensors can be used to determine the type of aircraft and the estimated speed of the aircraft passing over the inductive circuit sensor. The information about the type of aircraft and the position of the inductive circuit sensor can then be communicated to the central computer that merges the different data from the plurality of inductive circuit sensors and generates the tracking information about the aircraft or vehicles in the airport grounds. The central computer associates all the inductive circuit sensor data about a particular aircraft or vehicle since each inductive circuit generates a unique signal for each type of aircraft. Further details of the aircraft classification and tracking system according to the invention will now be described. Figure 3 is a block diagram illustrating more details of the aircraft tracking and classification system 40 according to the invention. The plurality of inductive circuit sensors 42 are connected to the electronic sensor components 44 located close to the runways and the raceways. The electronic sensor components can be partially powered in solar form and can be housed in a waterproof enclosure. The electronic sensor components 44 may include a detector 60, a classifier 62, a speed estimator 64, a communicator system 66 and an antenna 68 for transmng the generated data. The data generated by the electronic components of the sensor can also be transmd to the central computer through a wired communication line, such as a telephone line. As described below, various portions of the electronic components of the sensor can be implemented as software applications that are executed by a processor in the electronic sensors or as hardware circuits. The detector 60 detects the crossing of an aircraft or vehicle over one of the inductive circuit sensors and generates a signal when the crossing occurs. Classifier 62, based on the signal generated by the inductive circuit sensor, determines the type of aircraft for which the signal is most similar (ie, classifies the signal into an aircraft or vehicle type). The speed estimator 64, based on the signal from the inductive circuit sensor, determines the approximate speed of the aircraft or vehicle passing over the inductive circuit sensor. The communicator 66, which uses the antenna 68 or the wired communication link as described above, transmits the inductive circuit signal, the determined classification (ie, the type of aircraft), the velocity siphoned and any other information generated. by the inductive circuit sensor or the electronic components of the sensor to the central computer 48. The central computer, which may be located in the air traffic control tower may include the computer 50, the data fusion system / tracker 52 , the screen 54, an antenna 70, a database 72 and a record database of the aircraft 74. The central computer can also receive the data from the electronic components of the sensor on a wireline communication line. , such as a telephone line. The communicator 50, which uses the antenna 70, receives the data from the electronic components of the sensor 44 and corrects the data transmission errors, as described below. A portion of the information of the data received from the electronic components of the sensor can be new identification data from the electronic components of the sensor. The database 72 may be an identification database containing any type of known identifier data, any tracking data generated by the central computer and any other data relevant to the merging of the data or the tracking of the aircraft or vehicles, on the airport grounds. The data / tracked merging system 52 correlates the input classification data from the electronic components of the sensor to its aircraft or vehicle currently existing in data base 72 to determine if the data from that inductive circuit sensor correspond to an existing aircraft or vehicle. The data merger / tracker system may also update any tracking records based on the sensor data of the inductive input circuit and generate the data necessary to display the appropriate updated airport distribution as shown in Figure 1, in response to input inductive sensor data. Therefore, the screen 54 is constantly updated as new data is received from the inductive circuit sensor so that the air traffic controllers are constantly and automatically updated about the movement of aircraft and vehicles at the airport facilities. The distinctive registration database of aircraft 74, which is generated based on the data from the approach radar, allows the display to associate an aircraft flag which identifies the particular aircraft through its airline and flight number. , with the tracking information that is generated by the data fusion / tracker system. Therefore, the display can show an icon, with its distinctive information of the aircraft or vehicle, in the appropriate location of the aircraft or vehicle on the airport grounds. More details of the electronic components of the sensor will now be described. Figure 4 is a diagram illustrating more details of the electronic components of the sensor 44 and the circuit 42. The electronic components of the sensor 44 may include an analog-to-digital (A / D) converter 80, a waveform detection and wave extractor 82, an eigenvector decomposition system 84, a classifier 86, the speed estimator 64, the communicator 66 and the antenna 68. The electronic components of the sensor can be implemented, for example, as a plurality of software applications that are executed by a digital signal processor, such as a Motorola DSP 56002. The classifier 86 may be one of several types of classifiers, such as neural network classifiers of multilayer perceptron, of neural networks of basic or classic function. Each circuit 42 generates a variable inductance value. When nothing happens on a particular circuit, a stable or slightly variable inductance is measured. The slightly variable inductance when nothing happens on the circuit can be caused by the variations in the electrical components, due to the temperature, by variations in the moisture content of the floor in which the circuit is embedded, which causes the conductivity electrical field change, or variations in air temperature that cause the dielectric constant change. As described below with reference to Figure 5, the variation of the inductance can be accounted for by the use of an exponential filter. The detection software 82 detects an occurrence of a transient signal from an inductive circuit sensor that indicates the passage of an aircraft or a land vehicle over the inductive circuit. The transient signal is generated as the tip of an aircraft or the front of a vehicle passes over the circuit and continues to be generated until the tail of the aircraft or the rear of the vehicle has passed over the circuit. The inductance of the circuit is reduced when a vehicle or aircraft passes over the circuit. The transient signal will now be described with reference to Figures 5 and 6. The A / D converter 80 then converts the analog transient circuit signal into digital samples and the detection and waveform extractor 82 uses the digital samples to generate a digital signal. signal that is linearly related to the inductance of the circuit, as described below. The transient signal is detected as a change in the inductance from an average background value. The detection extractor and the waveform then extract a representation of a waveform from the time signal. To facilitate further processing of the signal, the temporal waveform is sampled again in a fixed number of samples, regardless of the time it takes for the aircraft or vehicle to pass through a particular circuit. The actual time required for the aircraft or vehicle to pass through the circuit is determined so that the speed of the aircraft or vehicle can be estimated by the speed estimator 64. The own vector decomposition system 84 and the classifier 86 use the form of Temporary wave to determine the type of the aircraft or vehicle based on the waveform, as described below with reference to Figure 5. The classification and speed estimation information for a particular junction of an inductive circuit can then transmitted to the central computer by any conventional communication system, such as an extended-spectrum radio communications system or a wired communications line. Now, they will be described in more detail about the detection and classification of the temporal waveform. Figures 5 and 6 illustrate more details about the detection extractor and waveform 82 and an example of the detected waveform respectively. As shown in Figure 5, the detection and waveform extractor may include a circular data buffer 90, a low pass filter 92, a second low pass filter, the interpolator and repetition sampler 94, a decision logic unit 96 and a temporary memory router 98. The signals representing the change in the inductance in the various inductive circuits are fed into the data buffer temporary 90. The low pass filter 92, to track the variations in the inductance of the circuits caused by non-events, such as changes in air temperature, changes in soil water contents and the like, as described above and generates a level of background inductance that is used as a threshold for the detection of an aircraft or real vehicle passing over an inductive circuit.

When a change in the inductance is the significant deviation from the background inductance level and lasts for a predetermined amount of time, this marks the beginning of something happening on an inductive circuit. The magnitude of the deviation sufficient to activate a crossover detection depends on the current value of the background inductance that changes as the inductance of the inductive circuit changes as described above. The decision logic 96 determines when the background threshold is exceeded and then the temporary memory router 98 stores the start address of the waveform once the threshold is exceeded. When the signal from the inductive circuit is below the threshold, the temporary memory router stores this address with the possible end of the waveform. If the signal remains below the threshold for a predetermined time, as T2, as shown in Figure 6, then a waveform end is indicated. On the other hand, if the signal again exceeds the background threshold by some time less than T2, then the end of the signal is not fixed until the signal remains below the threshold for at least the predetermined time T2. The waveforms extracted in this way have different durations depending on the speed of the aircraft or vehicle on the inductive circuit. In order to facilitate the classifications of these signals, each measured signal is sampled again by the repetition sampler 94 so that each signal output from the detection extractor and waveform 82 has the same number of samples. For example, in a preferred embodiment, each captured signal that exceeds the threshold is sampled again at a fixed number, z. of uniformly separated time samples and the duration of the original signal time, is saved in order to estimate the speed of the aircraft or vehicle. In addition to the aircraft or vehicle that passes completely over an inductance circuit, the system can also handle the situation where the aircraft or vehicle stops over the top of an inductive circuit. Pollo both, when an aero bird or vehicle stops on an inductive circuit the inductance of the inductive circuit decreases towards a lower value and remains constant as long as the object does not move. In this case, the detection and waveform extractor 82 detects that the aircraft or vehicle has stopped on the circuit, since the inductance is lower than the background threshold and has a constant value for a predetermined amount of time, such as several seconds. This information can be communicated to the central computer which then exhibits a representation of the vehicle or aircraft over the particular inductive circuit. An example of the signal captured by the detection extractor and waveform 82 will now be described. Figure 6 illustrates an example of the signal captured by the detection extractor and waveform according to the invention. The detection actions and waveform for a period from you to tlO, will now be described. Prior to time ti, the background threshold is tracked to determine the background threshold inductance signal. From it to t2, a time duration TI seconds occurs in which the input signal exceeds the limit determined by the current value of the background threshold inductance signal. Therefore, the probable start of something that passes over an inductive circuit is indicated and the direction of the temporary circular memory at time ti is stored as the probable start of detection. ? from the times t2 to t3, the change of the inductance signal is present and the circular temporal memory address in it is confirmed by indicating the start of the detection signal. From t3 to t4, a TI duration occurs when the input signal is below the threshold background inductance signal indicating the possible termination of the signal. The circular temporary memory address at t3 is tentatively stored as the end of the detection signal. From t5 to t6, the signal again exceeds the threshold of the background inductance. Since the period between the tentative end of the signal and the new signal exceeding the threshold is less than a predetermined amount of time T2, the new signals are part of the original detection signal and t3 is not the end of the signal of detection. From t8 to t9, the signal falls again below the threshold value and the circular temporary memory address at time t8 is saved as the possible end of the detection signal. One time duration T2 has passed without further signals exceeding the threshold so that t8 is confirmed at the end of the detection signal. This waveform, from you to t8 is passed to be sampled again, sorted and then the vehicle speed is estimated. Now, the process for classifying an aircraft or vehicle according to the invention will be described. The waveform generated, as described above, has a fixed sample number (z dimensions). In order to facilitate the classification process, the waveform can be reduced to a few samples (dimensions) through a decomposition of its own vector. as described below. Reducing the dimensionality of the data. the memory required to store the waveform is reduced and the number of inputs to the classifier is also reduced, which reduces the processing required for the classification of the waveform. Therefore, a simplified, smaller classifier, such as a neural network, can be used to process the waveform and a simplified classifier is faster, requires less memory to implement and runs faster than a more complex classifier. After the dimensional id of the waveform has been reduced, the classifier determines the classification of the aircraft or vehicle that is most likely to have produced the observed waveform and a clue of the most likely classification and the associated likelihood probabilities. they are then sent via the communications system to the data fusion / tracker system in the central computer. The reduction in size of the waveform will now be described in greater detail. To reduce the size of the waveform vector, a form of incipient component analysis is executed in which a two-dimensional matrix is formed by multiplying the waveform matrix of C, where each column of D is a fixed number of samples, M, the representation of the waveform. Then, D is a matrix I saw by N as large as data IDs has been >; an collected for, through its transposition, C = DD1. The normal result of this operation is a covariance matrix.

However, D was not normalized and the average was not subtracted from it. Next, C is subjected to the decomposition of the eigenvector which results in a set of M eigenvectors and associated eigenvalues. The eigenvalues constitute a set of orthonormal vectors that expand the M-dimensional space occupied by the input data. Therefore, all the input data can be decomposed in the sums of the eigenvectors. The eigenvalues associated with the eigenvectors give an indication that both of the variability in the original fixed number of sample waveforms can be projected onto the different eigenvectors. In a preferred embodiment, the largest eigenvectors where. K < M, can be used and provide sufficient resolution of the waveform detected to allow classification of the waveform. The classifier software can use a variety of different algorithms to classify the waveform. An example of a classifier is a modified elliptical base function (EBF) neural network that does a good job of classification and can be adapted quickly. It should be noted that other types of classifiers can also be used according to the invention. A neural network of elliptical basis function is an extension of a radial base function network (RBF), although the RBF forms spherical decision limits as opposed to the elliptical decision limits with the EBF. The RBF was originally created for biological systems in which neurons have selected responses for localized regions of the input space. In a conventional implementation, the inputs are fed in parallel to a set of neurons, each of which contains only a small portion of the total input space. The outputs of these localized neurons are weighted and added to a set of network output neurons. Figure 7 illustrates an example of a neural network of elliptical base function (EBF) 100 in which a plurality of input 1 2 are fed in parallel in a plurality of neurons 104 and the results from each of the neurons are fed in parallel to a plurality of output neurons 106. The input vector x, has K elements and the eigenvector is calculated from the identification identification of the aircraft. The function in the neurons 104. ((x.?, S,) is symmetric around each of the K axes where?, Is the mean of the own vector element ith for an individual class of aircraft and sr is the deviation standard of the own vector element ith for the same type of aircraft For the invention, the index i is between I and L since there are L different types of aircraft and vehicles that can be identified The values of?, and s, for each type of aircraft and vehicle are calculated from a database of identifications collected for each aircraft or vehicle, a greater number of real readings for a particular aircraft or vehicle, which constitutes a training of the neural network, increases the confidence that few aircraft of each type have their own vector output of 3s. An example of the actual implementation of the EBF classifier will now be provided for illustration purposes. \? Typical EBF is simplified to form a classifier. Or h i p e r i i t i c o by removing the pe in the outer layer. Each of the base function neurons 104 has a response for a region of the input space containing the entry starting from or at least of only a small number of aircraft classes or vehicles, depending on the way in which the input vector groups several classes of similar aircraft that can overlap. This can be determined through normal training procedures. The weights for a link between a base function neuron 104 and an output neuron 106 represents a class with a group center furthest from the center of the activation region that evolves to a small value. At the end of training, the output weights that have small values are removed which reduces the number of connections in the network and results in an output layer connected in a non-dense way. In order to develop this basic function classifier, it is assumed that each group has a Gaussian distribution when it is projected on each of its axes 1 <; and that the positions of the projections of the input vectors on the different axes are independent. The second assumption eliminates the need to store a covariance matrix K for K for each base function. These assumptions restrict the axes of the hyper-ipses that define the activation regions so that they are parallel to the axes of the signal vector itself, which increases the probability of the overlap of the activation regions for different classes of aircraft. Once sufficient data has been collected, the values of vectors?, And cj, can be obtained for each class of aircraft or vehicle. When an input vector to the classifier is presented, the square of a type of hyperelipsoidal distances Y, between this input vector and each of the group centers is calculated c o ni o: From the assumption of independent Gaussian distributions for the different values of x, it is assumed that Yj has a chi square probability density function for n = K degrees of freedom so that: ^%? -%? 0 = Á n An) ar each base func tion, and calculates a value Y then integrates the previous density function from Y; until infinity. This provides a number that can be considered as a probability that, given an input vector representing a class j, the calculated value of Yj would be equal to or greater than Yj. This is then the activation function for each of the neurons of base fusion: K (x,? ¡, s = P (Y> Yj / H) = P (Bj / Hf) Figure 8 shows a graph of the previous activation function It is possible to numerically integrate the appropriate function for calculate K (x.?,. o-,) for each base function neuron each time an input vector is presented to be classified, which requires some extra processing code and calculation time. visualization of the activation function with the hyperbolic tangent function that is normally used for the front-feed multilayer perceptron activation functions, it has been found that an approximation to K (x,? ,, s,) could be obtained using only two hyperbolic tangents that have a half-squared discrepancy with (x,? ¡, s¡) of only .002 over the range of 0 to 50. This approach has the advantage of small code space to implement and less computation time. Therefore, to calculate the probability that the aircraft or vehicle that generated the waveform is of class j, given ¡from the base function neuron j, using the Bayes theorem so that: The values of P (Hj) are estimated from observed frequencies of the different classes of aircraft or vehicles. One of those values for each possible class is necessary. Any attempt to measure the reliability of this injection vector is necessary because the waveform could have been generated by an unusual type of aircraft or possibly a land vehicle for which classification has not been encoded, or any of the many possible events that could have happened to produce some distortion in the waveform. A measure of certainty of the Pearl tree injection vector is also necessary. The above process can generate a reasonable injection vector even if the maximum value of P (ßj / H.}.) Is small. Therefore, we use the largest value of P (ßj / Hj) as a measure of certainty. Once the most likely classification of the aircraft? vehicle has been determined, the speed estimator 64 estimates the speed of the object. In particular, when an aircraft has been classified, its length can be determined using a search table. Then, the length of the aircraft is divided by the time required for the aircraft to pass over the circuit to determine an estimated speed in feet per second that can be converted to knots. The communication system 66 then transmits any data to the central computer 48. The communications system handles the loss of data that may occur. One of the electronic components of the sensor 44 can be placed around an airport so that there is a total of, for example, four communication zones that have some overlap coverage, although with faster communications since there are only four zones. One zone represents an individual wired communication line or radio channel depending on whether or not a wired communications system or a spread spectrum radio communication system is used, respectively, to communicate the data between the sensor electronic components and the central computer. Each of the electrical components of the sensor 44 can be registered once in each predetermined amount of time, such as once per second. During verification, the central computer will request any detection data. If there is no detection data, a "no data" message is sent to the central computer. If the detection data exists, then the identification waveform, its classification, its speed estimation and any other data are sent to the central computer. An example of the data fields that can be sent to the central server are listed in the following table. Table 1 Data Field Name Size Description Loop_ID 1 Cycle Number Date 4 Date and time identi cation was collected Acceleralion 4 Estimated acceleration of the aircraft or vehicle S eed 4 Estimated velocity of the on eve eve w orch 1 or I nterva 1 2 Amplitude of time of each sample of identification that was re-sampled. Max I nductance 2 I nductance maximal detection Min Inducing ce 2 I nductancia mimima of detection OriginalJ ength 2 Original length of identification before the new model Signature 64 Identification vector No C 1 ass 1 Number of classifications in Classifi cation 4 Classi fi cation, size 10 Likelihood 4 Probability of correct classi fi cation, size 10 For, each message sent to the central computer. a checksum is calculated by the electronic components of the sensor, and appended to the end of the message so that the central computer can use the checksum to verify the message for errors. If the control sum calculated by the central computer is different from the control sum at the end of the message, the central computer requests the retransmission of the last message. Similarly, for messages from the central computer to each of the electronic components of the sensor, the central computer calculates a sum of control and the adjoining one at the end of the message so that each of the electronic components of the sensor can verify the message for errors and request retransmission as necessary. Any information about the non-operating electronic sensor components or circuit can be passed to the central computer because the central computer periodically consults one of the electronic components to make sure that the circuit is functioning properly. Otherwise, the electronic components of the sensor may send a "bad circuit" message back to the central computer. The details of the central computer will now be described more thoroughly. Figure 9 is a diagram illustrating the details of the central computer system 48 according to the invention. The central computer 48 may include the communications system 50, the data fusion / tracking system 52 and the screen 54. In addition, the host may also include a database 120 and a temporary memory 122 connected to the communications system 50. , an interprocess communication system 124 which may be a connector or some other communication system, which connects the communication system to the data fusion system / tracker, a tracking database 125 connected to the data fusion system / tracker, second interprocess communication system 126, which may be a semaphore, a combination of shared memory (SHM) and a connector or some other communication system, connected between the data merger system / tracker and the screen, and a flag tracking database 1 2 connected to screen 54. Communications system 50 can be implemented as a software application that is executed by a processor in the central computer. The communications system can have five processes. In this example. there can be a communication process 134-140 for each of the communications zones in the airport and a communication control process 142 that receives the data from each of the zone processes. Each zone process 134-140 can receive data from the inductive circuits and the electronic sensor components within the communications zone. The communication system 50 according to the invention can have any number of corresponding zone and communication zones depending on the size of the airport. The communication control process 142 initiates the zone communication processes to register each electronic sensor component in a predetermined range to test the operation of the electronic sensor components and the inductive circuit sensing. The communication control process may also collect the detection data from the zone communication process for the current verification step and initiate the zone communication processes for the verification of their respective electronic sensor components again. The verification of the electronic sensor components in each zone can be a continuous process controlled by the communication control process. The communication control process writes the identification and classification detection data for the database 120 for future use and sends the detection data of the aircraft and the land vehicle to the Data Fusion / Tracker system 52. The different data detection which can be communicated between the central computer 48 and each of the electronic components 44 was described beforehand. Each central computer zone process 134-140 communicates with the electronic components of the sensor 44 assigned to the zone and, in each communication with one of the electronic sensor components, requests any detection data. In case no aircraft or land vehicles have passed over the circuits connected to the electronic sensor components, then a "no data" message is sent back to the central computer. If the detection data is available, your identification, classification, speed estimation and other associated data are sent to the central computer. The communication control process 142 writes the detection data to the database 120. The input data may be temporarily stored in the temporary memory 122 before storage in the database 120. Table 2 below identifies an example of the different data field definitions for the data records of the database 120 in which all the identification data and classification of aircraft are stored.

Table 2 Database Data Fields Name Size Description Loop D 1 Circuit Number Date 4 Identification of collected date and time Accel erat ion 4 Estimated Acceleration Speed 4 Estimated Speed I nterval A plitude of time of each newly sampled identification sample Max Inductance of Maximum Decentrality Inductance M in I nductance Minimum Detection Attribute Original Length Original Identification Length before the new sampling Signature 64 Identification vector Num Class 1 Number of c i a s i f i c a n c e s in the track Classification Probable classification. size 10 L i k and I i h o o d Probability of correct classification. size 10 IVS-2000_Freq Frequency of electronic sensor components used to boost the Sample_Rate circuit Speed of the electronic component of the sensor Flight Num Distinctive number of the Aircraft TNuin flight 10 Tailgate number T ruth C 1 ass Classification of aircraft rea 1 T r LI lh S peed Aircraft speed rea 1 Temperature 2 T emperature Wind_Vel 2 Wind speed Wind_Dir 2 Direction of wind Baro Press 2 P a tion Ba mét ica After the communication control process 142 collects the detection data from each of the communication processes of zone 134-140, the deletion data are classified, in time of the order of detection, and sent to the Dalos / Tracker Fusion system 52. Table 3 illustrates an example of the definition of the data interface between communication control processes. and the Data Fusion / Tracker system 52.

Table 3-Data Fields Between the Communication Process and the Data Fusion System / Tracker and the Screen ombp T ama ñ o Description Loop ID 4 Nedumber of the peed circuit F 1 ag 4 1 ndicasi 1 at speed estimable of the classifier is Speed 4 Speed and detection Num C 1 ass 4 Classification number following the Type 4 ('1 classifications up to 10 Class Prob 4 Classification probability, up to 10 Signature 4 ID Identification Identification Sizes 64 Detection Time 8 Detection of time that was made by the Electronic Components of the Sensor Any detection data is sent to the Data Fusion / Tracker process 52 through the communication process 142 through the interprocess communications 124. which may be a connection. There may be six functions that are implemented to handle all interprocess communications between the communications process 142 and the merge process of the / trace 52. These functions may include for example • initializing interface input: Called by the communications process and the Data Merger / Tracker process to initialize the interface and assign a file descriptor for the Inlerprocess Communications. • Enter message type: Called by the communication process or by the data / tracker merge process to determine the particular type of pending message from the other processes.

• Send message entry: Called by the communication process or by the data merger / tracker process to send a message to the other processes. • Receive message input: Called by the communication process or by the data / tracker fusion process to retrieve the message from the other processes. • Enter descriptor file: Called by the eomunicaicons process or the data merger / tracker process to obtain the interface file descriptor for that process to use in the interprocess communication connection. • Interface input shutdown: Called by the communications process and the data merger / tracker process to deactivate and close the interface between the communications process and the data merger / tracker process. »Send track tracing: Called by the process Data Merger / Tracker to send a trace track to the free temporary memory of the shared memory interface with double temporal emotion. The data fusion / tracker system 52 is now described, which may be a software application that is executed by a process of a central computer 48. In a p r e f e r e d a r e d a mode. the data fusion / tracker process can be a multiple hypothesis tracker software application (modified to process the detections of aircraft and land vehicle.) The data merger process / tracker 52 accepts the detection inputs from the process of communications 142, form traces and link new lessons to the existing traces and send the tracking information to the palette 54. The screen 54 is updated periodically and preferably about once per second.The refresh rate of the screen is a modifiable parameter that can be established at the start of the system Figure 10 illustrates an architecture for the data merging / tracing process 52 according to the invention The data merge / trace process can include an aircraft configuration file (FIG. ACF) 150, a communication interface 152, a tracking partner 154, a kinematic switch 156 and a sorting switch 158 connected to the tracking partner, a tracking updater 160, a kinematic updater 162 and a sorting updater 164 connected to the tracking updater, a tracking store 166, a tracking predictor 168. the security logic 169 and the screen interface 170. The data merging process / tracker 152 is automatically configured by reading the airport configuration file (ACF) 150. This file contains information about the configuration of each airport in which the system is installed. For example, the file may contain information about the Long Beach Airport. The other functionality within the data merger / tracker process is independent of the airport and it is only the ACF that needs to be modified when the system is installed at other airports. The airport configuration file (ACF) 150 is read by the data merger / tracker process when it starts to operate. The ACF defines runways and runways in terms of segment. The segments are defined as those sections of the runways and runways that lie between the intersections of the runways and the runways. The ACF can also define the transition rules to transfer from one segment to another within the airport. The ACF also defines the placement of the inductive circuit sensor at the airport and the distances between each of the inductive circuit sensors. As described above, the other data fusion / trace functionality is independent of any specific airport configuration file. The data fusion / scanner 152 for the communication process is triggered by event. In other words, when the detection of an aircraft occurs and it is received by the central computer, the data merging process / crawler is automatically notified by the communication process through the communication interface 152 which may be a connector interface. on UNIX basis. The trace association process 154 runs the test to determine if any inbound detections can be linked to the existing crawls. Otherwise, a new trace is created. The trace association process 154 uses the definitions in the ACF to apply the circuit transition tests. Two types of selection of the new detections with the existing traces were made: The classification section 158 and the dynamic selection 156. The classification selection, for association, uses the Bayesian Pearl tree. In particular, based on the classification data from the detection of the electronic components of the sensor, the probabilities of a particular tracking that is of a certain aircraft are updated by the Bayesian Pearl tree. The kinematic selection 156, on the other hand, uses the Kalman filter innovations. The classification and ordering of association probabilities are executed to determine the final discrimination for association of a new detection for a tracking. The 160 trace update is executed using the one-dimensional Kalman filter estimates to determine the tracking position and velocity and the Bayesian Pearl tree to integrate, and thus improve, a tracking classification of the aircraft. The trace update process 160 also calculates the circuit-to-circuit acceleration metric, the circuit forecast time error metric and the classification refinement metric (improvement of classification security from the detection of the first circuit ). The tracking forecast 168 can determine the forward forecast of tracking over time and can be achieved by applying a Kalman filter forecast between tracking updates. The decision logic in this function uses the ACF to control the tracking forecast at the intersections of the runway and runway and on the circuits. The data merger / tracker process for the screen interface 170 can be implemented using interprocess communications, such as a shared memory and a UNIX connector interface. After any crawls have been updated. All traces are forecasted to go to the current time and sent to the exhibition process. The tracking forecast process 168 may occur once in each predetermined interval, like once per second, without considering whether or not the detection has been presented. This is to present the user with a uniform movement tracking and to avoid having tracking positions that change from circuit to circuit. The "downstream tracking" and "downhill circuit" operator commands are sent to the data fusion / tracker 52a process from the screen by means of data / trace fusion for the screen interprocess communications. An example of the detection data interface and the data fields communicated between the data merger process / tracker and the screen run are shown in Table 4. Table 4-Data Fields Among the Data Merge Process / Tracker and Display Name T ack Description T ack State Time 8 Tracking position time Trk_Display_ID 4 Tracking number Last Loop ID 4 Last detected trace circuit X .Offset Position coordinate X coordinate And Offset - Position screen Y coordinate Pre_Track Flau Indicates first trace detection, can not start tracking only with a detection Course 4 Trace course Speed 4 Tracking speed Delta_Class Prob 4 Improvement of the probability of classification from the first detection Loop_Predic_T¡ me 4 Error of time of the trace that is detected in the following circuit Num Class Number of classifications that follow Class_Type Classification type, up to 10 Class Prob Classification probability, up to 10 Signature Identification Vector Size 64 Any detection data is sent to the Screen process through the Data Merger / Tracker process through a communication system of i n t e r p r o c e s t t as a shared memory. In the following example, four can be implemented to handle all communications between the Data Merger / Tracker process and the screen process, although there may be additional functions according to the invention. • Initialize interface screen: Call by the Data Fusion / Tracker process and the Screen process to initialize a shared memory interface with double temporal memory between the Data Merge / Tracker and the Screen process. • send trace track: Called by the Data Merge / Tracker process to send a trace track to the free temporary memory of the shared memory interface with dual temporary memory. • receive track trace: Call by the Screen process to verify the temporary memory of the free shared memory for a new track trace from the Data Merge / Tracker process. If a new tracking track is available in the free temporary memory, the return value is set to one and the pointer of the tracking track is set at a point for the temporary memory with a new track trace. The screen process has temporary memory and the data merger / tracker process can not write to that temporary memory until the screen process requests the next trace track. A return to zero indicates that the data merge process / crawler does not have a new trace track yet. • Turn off interface screen: Called by the data merger / tracker process and the screen process to finish the interface between the data merger process / tracker and the screen process. The invention can use different types of interprocess communications, such as shared memory and a coneclor. The shared memory process communication can use the four processes described above, while the connector can use up to 6 functions, as described above, to handle communications to the data merger process / tracker. In one example, shared memory communications handle a majority of the data between processes while the connector passes the descent and tracked descent circuit commands. The Data Fusion / Tracker process 52 may receive a "downlink" message from the Communications process or the Screen process. Yes. At any time, an inductive circuit sensor becomes inoperative or the electronic components of a sensor become inoperable, that information must be passed to the data merger / tracer process. The central computer periodically queries each of the electronic sensor components to ensure that its circuit is functioning. If a circuit is not working, the electronic components of the sensor send a "bad circuit" message to the central computer. The central computer then sends a m e n of "down-cycle" to the data merger / tracker process. If the electronic components of the sensor do not respond to the query, the central computer sends a "downlink" message to the data merger / tracker process. The screen process due to the input from the operator, can send a message of "descent circuit" to the data merger / tracker process, ready it can happen due to some change in the operation characteristics of the circuit. If the detection identifications that are being received have been significantly changed and the classification process is of an incorrectly classified aircraft, then it is best to omit that circuit from the additional processing.

There are two methods by which a crawl can be lowered by the data merger / crawler process. With any method that is invoked, that trace is descended from the additional processing by the data merger / crawler process. First, if an existing trace is not updated for a fixed amount of time. the data merger / crawler process will automatically delete the crawl. The time parameter is set in a parameter file and is editable. An automatic tracking elimination may occur because the airport is not fully instrumented in circuits. Therefore, aircraft and land vehicles can enter and leave the instrumented portion with airport circuits at many points from other runways or runways that do not have inductive circuit sensors installed. In order not to obstruct the screen with tracking, those traces are automatically deleted. The aircraft that has taken off has its traces automatically removed by the tracking update interval feature. Secondly, the operator has the option to download a trace from the data merger / crawler process without the automatic crawl removal time is greater than desired. This provides a method for the operator to clean the screens. The screen is used to display the traces of aircraft and land vehicles at the airport, such as the Long Beach airport. The screen process can be implemented as a software application. The screen receives the tracking information from the data merger / crawler process through the shared memory interface between these and then displays the updates for the trace. Each trace has its tracking number displayed next to it. The data merger / crawler process may also include some raid logic that can be implemented in the software within the central computer. This is modeled in the data fusion architecture of Figure 10 as security logic 169. The raid logic can warn an air traffic controller of the tower when there is an unsafe condition on the take-off or runway tracks of the tower. airport. An incursion has been defined by the Federal Aviation Administration (FAA) as any occurrence at the airport controlled by a tower that involves an aircraft, vehicle, person or object on the airport ground that creates a coalition risk or results in loss of separation with the landing or takeoff of aircraft. The display system, as described above, can display a predetermined number of different alarms for example, in order of priority ascension and continues to display the weapon's messages until the raid conditions disappear, the alarm displaces the screen due to High priority messages or alarm 1 cleans the alarm. The alarm can be a warning message that flashes intermittently or caution messages that appear on the screen in which a warning message is of greater priority interest than a caution message. For both warnings and precautions, an audible alarm may sound for some predetermined period, such as three seconds. For a caution alert, a single short, high-pitched sound may sound.

The system operator may be able to customize the different warning and caution indicators, such as adjusting the volume or length of the audible signals. In the raid logic, each event is detected and can be classified within several priority levels in which the warning messages have a higher priority than the caution messages. Therefore, for multiple events, events are organized on the screen based on the priority level associated with the event. For example, events in which there are two traces (ie two different aircraft or vehicles) may have a higher priority than events in which there is only one trace (ie a single event involving only one aircraft or vehicle). As another example, an event in which two aircraft approach each other (ie, an approach geometry), can be assigned a higher priority than an event in which two aircraft are separating from each other ( that is to say, a geometry of to 1 ejamiento). As another example, events involving the departure, arrival and / or landing of the aircraft receive priority over events involving the bearing and / or stopping of the aircraft. The raid logic can decept several different events / raids that may include but not be limited to: 1) frontal traffic in which two aircraft move toward each other in opposite directions; 2) heading / landing in the opposite direction in which the aircraft is moving downward in a takeoff view in a direction opposite to the expected direction; 3) multiple exits in which a departing aircraft is followed very closely by another aircraft departing on the same take-off runway; 4) Exits on a busy take-off runway where it is assumed that an aircraft is going to leave on a runway which is already occupied by another aircraft that may be stopped threate the take-off. landing or bearing; 5) the take-off runway is occupied by an aircraft that is stopped; 6) two aircraft on crossed runways where the aircraft can enter a runway intersection at the same time; 7) two aircraft are rolling in the opposite direction on the same bearing track in which an aircraft does not have an exit from the raceway; 8) an aircraft arrival on a busy take-off runway; 9) frontal traffic in which one aircraft is arriving and the other aircraft is moving towards the arriving aircraft; and 10) with a properly located circuit sensor, an alert may be provided if an aircraft or land vehicle enters an area of the airport that is not allowed to enter. Now, an example of the trace symbol for a small aircraft, its tracking number and its identification analysis is described. Figure 11 shows an example of a tracking symbol for a small aircraft, its tracking number and the identification analysis window. The tracking and aircraft information 58 can be displayed by selecting the operator of several options. The identifications of aircraft or land vehicles that pass over any circuit can be brought to the screen for examination and comparison with their identification from another circuit. An identification window 160 contains the information about that associated detection and tracking information. The information displayed about the circuit detection includes the circuit, number of detection, c 1 a s i f i c a n e a on, security in the classification, speed and course. The operator has the option to eliminate a circuit or trace from the additional processing. The screen process due to the input from the operator, can send a message of "descent circuit" towards the process of fusion of dalos / tracker. This may occur due to some change in the operating characteristics of the circuit. If the detection identifications that are received have changed significantly and the classiication process is misclassifying the aircraft, then it is better to eliminate that circuit from further processing. When the operator selects a trace to be downloaded from the screen. A message is sent to the data fusion / tracker process. The data merger / tracker process then descends that trace from the additional processing. The detection data can be stored in a database and then consulted based on different criteria. The detection data can be extracted by date, aircraft type, flight number, tail number, and badge number. For example, several scenarios can be created by extracting the identifications for a particular aircraft for a specific day. These scenarios can be created for tracking and classification purposes or for demonstrations. The scenarios created in this way are processed through the data merger / tracker process and the screen. They can also be processed through the operation of the classified in the central computer, since there is no method of feeding the stored identification and the detection data through the components and reading them from a sensor. This is the method used to train the classified and the Performance evaluation. The additional analysis can be performed to identify the individual circuit differences and characterize the differences in identification by circuit size. Another use for the archived data is to perform the identification analysis of the same aircraft over time in an attempt to identify and characterize any changes in the circuit or deterioration of the electronic components over time. The system may also include performance analysis systems. The performance analysis measures how precisely the system behaves. A number of metrics have been defined to measure this performance. The metrics are divided into two classes: detection and classification metrics and tracker metrics. An example of these metrics is described below. • Detection and Classification Metrics 1. Correct Classification: This metric measures how capable the classifier is of classifying an aircraft according to the size, small, large or heavy, compared to the total number of correctly classified aircraft for the number of aircraft that was visible for classification. This number includes those aircraft that were not classified. The correct classification number and percentage will be reported.

Incorrect Classification: This line measures how poorly the classifier behaves according to the aircraft size, small, large or heavy, by comparing the total number of aircraft incorrectly classified for the aircraft number that was visible for classification. The number visible to the aircraft is included in aircraft that were not classified. This number is the inverse of metric 1, previous, minus those aircraft not classifications. The number and percentage of the incorrect classification will be reported. Unmetered: This metric measures the number of the aircraft that the classifier was not able to classify when comparing the total number of aircraft not classified for the aircraft number that was visible for the classification. The number available for classification includes those aircraft that were not classified. The number and percentage of aircraft not classified or classified will be reported. Cycle Crossings for Correct Classification: This metric measures how quickly the classifier is able to correctly classify an aircraft by measuring how many circuits it must pass. an aircraft before it is correctly classified. The average deviation and the standard deviation of the number of circuits it takes to classify an aircraft correctly will be reported. Those aircraft not classified correctly will be excluded in this metric. Correct Classification by Specific Aircraft Type: This metric measures how well the classifier can classify an aircraft according to the type by comparing the total number of correctly classified aircraft for the number of aircraft that were available for classification. The number available for classification includes those aircraft that were not classified. The number and percentage of the correct classification by the specific aircraft type will be reported. Incorrect Classification by Type of Aircraft E s p e c i f i c a: This metric measures how well the flight classifier behaves with the type of aircraft by comparing the total number of incorrectly classified aircraft with the number of aircraft that were available for the aircraft. classification. The number available for classification includes those aircraft that were not classi fi ed. This number is the inverse of metric 5 above, minus those unclassified aircraft. The number and percentage of incorrect classification by specific aircraft type will be reported. Correction of Speed Estimation: This metric measures what is capable of estimating the speed of the aircraft correctly. The metric is reported as a mean difference from the current speed and standard deviation. Matrices and Trackers 1. The number of circuit detections for tracking to bring them to the screen: This metric reports how many circuit detections it takes on average for the tracking of an aircraft to be displayed. They are reported in the mean deviation and the standard deviation. 2. Probability of correct classification (Pee) for the first circuit against current circuit detection: This metric is used as an indication of how the number of step circuits of an effective aircraft improves the correct classification. This metric will be reported as a table indicating the average and standard deviation of the improvement from the first circuit to the second circuit, the first circuit to the third circuit, the first circuit to the fourth circuit, etc. 3. Predicted circuit detection time of an aircraft tracker against the actual detection time: This metric is generated by the data merger / tracker process, which predicts for each tracking the time of arrival in the next circuit . This metric reports the average and standard deviation of the differences between the arrival time of the circuit and the actual circuit detection time as reported by the IVS-2000. 4. False traces: This metric reports the number and percentages of traces that were incorrectly formed by the data merger / tracker process 5. Lost traces: this metric reports the number and percentages of detections that were not sent to the data merger / tracker process . As described above, the detection scenarios for analysis can be generated by a selection of aircraft series by number of badge or number of queue for a given day from the database. These scenarios are processed through data merger / tracker and performance statistics are collected. These statistics are used to identify classification problems and tracer deficiencies. If classification problems are found, the training and additional classifier may be indicated or an increase in the number of standard deviations allowed about the mean for any identification vector may be required. _ Three demonstration modes that offer different levels of operator interaction and control are available.

For the real-time processing mode and the type of scenario demonstration generated from the database, the operator has the ability to eliminate the circuits and track from the screen, causing additional processing of the deleted and traced circuits until the detection . For tracker output file processing, no operator interaction is allowed. This is just a mode of reproduction.

The real-time processing demonstration mode uses 1 software described above. The electronic sensor components collect the data and that detection data is sent to the central compiler. The data / crawler merge process generates and updates the traces based on the aircraft or land vehicle detection data received by the electronic components of the sensor. These traces are sent to the screen where they are displayed. Total circuit operator control and tracking eliminations are available in this demonstration mode. Using the detection scenarios created from the aircraft identification database, a demonstration of any length can be created. This data is read by an input process that replaces the communication process and sends the detection data to the data / tracker merge process. The data merger / crawler process generates and updates the traces based on the aircraft or land vehicle detection data received from the scenario. These traces are sent to the screen where they are displayed. Total circuit operator control and tracking eliminations is available in this demonstration mode. I For the demonstration of output file processing of the tracker, the screen process receives the data from a program that reads a saved tracker output file. For this demonstration mode, no screen operator interaction is available. It only displays those traces that were previously generated by the data merger / tracker process.

While the foregoing has been with reference to a particular embodiment of the invention, those skilled in the art will appreciate that changes can be made in this embodiment without departing from the principles and spirit of the invention, the scope of which is inido by the appended claims.

Claims (12)

1. CLAIMS 1. A system for classifying and tracking traffic at an airport that has at least one runway and at least one runway, the system includes one or more sensors placed at predetermined locations near the take-off runway and the bearing track for generating a signal in a vehicle passing close to a particular sensor characterized in that: the signal is produced by the particular sensor in response to the vehicle passing close to the particular sensor and that the system also includes: means for assigning a classification to the vehicle in response to the signal; means for determining a vehicle position at the airport as the vehicle passes close to one or more sensors; and means for displaying a representation of the vehicle at the airport in the representation of the vehicle indicating the classification of the vehicle and the position thereof.
2. 2. The system of c or n f or r m i d a d with claim 1, characterized in that the classification of the vehicle is generated from the interaction between the vehicle and a magnetic field created by the sensor.
3. 3. The system according to claim 1, characterized in that the classification of the vehicle is generated from a decrease in an inductance of the sensor.
4. The system according to claim i, characterized in that the classification of the vehicle is generated by monitoring a signal from one or more sensors as the vehicle passes close to one or more sensors and comparing the signal with a plurality of sensors. vehicle identifications, each vehicle identification that is associated with a different vehicle classification.
5. 5. The system according to claim 1, characterized in that the classification of the vehicle is generated without receiving a signal that is transmitted from the vehicle and that indicates the classiication of the aircraft.
6. The system according to claim 1, characterized in that each sensor comprises an inductive circuit sensor.
7. 7. The system according to claim 1. characterized in that each sensor comprises an inductive circuit located below the runway or the runway.
8. The system according to claim 7, characterized in that the inductive circuits are embedded in the material used to make the runway or the runway.
9. The system according to claim 1, characterized in that the display means comprise means for generating and displaying an exit track.
10. The system according to claim 7, characterized in that it also comprises means for calculating the speed of the vehicle passing over an inductive circuit.
11. 11. The control system with claim 1. further comprising the electronic sensor components located in a waterproof housing near the take-off runway and the bearing track.
12. 12. The system in accordance with the claim 11, characterized in that the electronic sensor components further comprise a detector for detecting the crossing of a sensor and generating a detection signal, a classifier for selecting a vehicle identification from a predetermined number of vehicle identifications based on the detected signal, means for estimating vehicle speed based on the detected signal, and means for communicating information about the vehicle to a remote central computer. 1 . The system of c o n f o r m i d a d with the claim

    12. characterized in that the classifier comprises a neural network. The system according to claim 1, characterized in that it also comprises means for tracking data to determine if the signal is equal to the already existing signals. 15. The system of c o n f o r m i d a d with l vindication 14, characterized in that it further comprises a database of known identification values and means for comparing the input signals to the database to determine if the input signal is related to the same vehicle. 16. The system in accordance with the claim 15, characterized in that it also comprises means for generating an indication of the movement of the vehicle around the airport based on the signal to generate a tracking of the vehicle. 17. The system in accordance with the claim 16, characterized in that it further comprises means for selecting the tracking of the signal to an existing trace if the input signal equals an identification of an existing trace. 18. The system according to claim 1, characterized in that it further comprises a database containing the aircraft registration information to match the input signal to a flight number. The system according to claim 1, characterized in that it also comprises means for detecting a stopped vehicle on the sensor. 20. The system according to claim 1, characterized in that the vehicle comprises an aircraft. 21. The system according to claim 1, characterized in that the vehicle comprises a land transportation vehicle. 22. The system according to claim 1, characterized in that it also comprises means for determining an incursion between two vehicles on the airport grounds based on the signal for each vehicle.