Intelligent inspection method for park scenic spots based on environmental sanitation integration

Abstract

The present application relates to the field of electronic digital processing technology, and specifically to an intelligent inspection method for park scenic spots based on integrated sanitation, the method comprising: calculating equipment abnormal values ​​of different sanitation equipment through historical inspection data; calculating predicted abnormal values ​​through detection data change characteristics of different sanitation equipment in combination with inspection time; constructing an inspection plan vector, calculating an inspection plan evaluation value according to the impact of the inspection plan on the equipment abnormal values ​​of the entire inspection area, and selecting the best inspection plan in combination with a genetic algorithm; guiding inspection personnel to complete the inspection according to the selected best inspection plan, and compared with the traditional fixed inspection plan, the method can dynamically perform equipment inspection according to the situation of the sanitation equipment, avoiding the problem of waste of inspection resources and blind inspection angles due to improper allocation of equipment inspection resources.

Description

Technical Field

[0001] The present application relates to the field of electronic digital processing technology, and specifically to an intelligent inspection method for park scenic spots based on integrated environmental sanitation. Background Art

[0002] Environmental sanitation is crucial to the development of park scenic spots. Good sanitation in park scenic spots can increase the number of visitors to the scenic spots and directly or indirectly affect the development of the surrounding economy, society and culture. However, before ensuring the sanitation work in park scenic spots, it is necessary to ensure the full maintenance and management of the sanitation facilities in the park scenic spots to ensure that the sanitation facilities can be used normally when needed. Therefore, it is necessary to fully inspect and maintain various sanitation facilities in park scenic spots.

[0003] Existing parks and scenic spots mostly use regular inspections to inspect and maintain equipment, but fail to make full use of the intelligent information of the equipment. When a piece of sanitation equipment is repeatedly used at a high frequency in a short period of time, the urgency of the maintenance of the sanitation equipment increases rapidly. At this time, the use of a fixed inspection plan will cause this part of the sanitation equipment to be in a blind spot in the inspection. When some sanitation equipment is not used for a short period of time, but these sanitation equipment need to be inspected according to the fixed inspection plan, there will be a waste of inspection resources. Therefore, in the current sanitation inspection process of park scenic spots, there is a situation where equipment inspection resources are improperly allocated due to improper inspection planning, which leads to the problem of easy waste of inspection resources and inspection blind spots. Summary of the invention

[0004] In order to solve the above technical problems, the present application provides an intelligent inspection method for park scenic spots based on integrated sanitation to solve the existing problems.

[0005] The intelligent inspection method for park scenic spots based on sanitation integration of this application adopts the following technical solutions:

[0006] An embodiment of the present application provides a park scenic area intelligent inspection method based on environmental sanitation integration, the method comprising the following steps:

[0007] Collect various inspection information of each type of sanitation equipment, including attribute information and time information of sanitation equipment;

[0008] Based on the inspection information of each sanitation equipment at each inspection in the historical time, each historical detection vector of each sanitation equipment is constructed; the inspection information of each sanitation equipment at the current moment is obtained to construct the real-time detection vector of each sanitation equipment; the remaining sanitation equipment except the sanitation equipment inspected at the current moment is recorded as pending equipment; based on the difference between the real-time detection vector of each pending equipment and the historical detection vector of the same sanitation equipment, as well as the maintenance status of the same sanitation equipment at each inspection, the equipment abnormal value of each pending equipment is constructed;

[0009] Based on the variation between the same type of elements in the historical detection vectors of any two sanitation equipment of the same type, each sample time variation vector and each sample feature variation vector of each type of sanitation equipment are constructed. The feature variation vector of any pending equipment is predicted by combining the time required for the sanitation equipment inspected at the current moment to reach any pending equipment and the time information of any pending equipment at the current moment. Based on the real-time detection vector and the predicted feature variation vector of any pending equipment, a predicted detection vector of any pending equipment is constructed, and the predicted abnormal value of any pending equipment is obtained by calculating the abnormal value of the equipment.

[0010] A preset number of inspection plans are randomly generated, and the inspection plan evaluation value of each inspection plan is calculated based on the equipment abnormal value and predicted abnormal value of the pending equipment, so as to obtain the best inspection plan for sanitation equipment inspection.

[0011] In one embodiment, the time information is:

[0012] The time difference between each inspection time of each sanitation equipment and the corresponding last maintenance time is recorded as the maintenance interval time of each sanitation equipment each time it is inspected; and the time difference between each inspection time of each sanitation equipment and the corresponding last inspection time.

[0013] In one embodiment, each historical detection vector of each sanitation equipment is a vector composed of all attribute information and all time information of each sanitation equipment each time it is inspected.

[0014] In one embodiment, the process of acquiring the real-time detection vector of each sanitation equipment is as follows:

[0015] The attribute information of each sanitation equipment at the current moment is used as the real-time attribute information of each sanitation equipment; the time interval between the current moment and the time when each sanitation equipment was last inspected is used as the real-time inspection interval of each sanitation equipment; the time interval between the current moment and the time when each sanitation equipment was last repaired is used as the real-time maintenance interval of each sanitation equipment;

[0016] The vector composed of the real-time attribute information, real-time inspection interval and real-time maintenance interval of each sanitation equipment is used as the real-time detection vector of each sanitation equipment.

[0017] In one embodiment, the process of obtaining the device abnormality value of each pending device is as follows:

[0018] If the sanitation equipment is repaired each time it is inspected, the tag value of the sanitation equipment is set to 1 each time it is inspected, otherwise it is set to 0;

[0019] Calculate the Euclidean distance between the real-time detection vector of each pending device and each historical detection vector of each sanitation device of the same type, and record it as the first Euclidean distance; use the normalized value of each of the first Euclidean distances of each pending device as the distance weight of the corresponding historical detection vector, and record the product of the label value corresponding to each historical detection vector of each pending device and the corresponding distance weight as the first product; and use the sum of the first products of all historical detection vectors of each pending device as the device abnormality value of each pending device.

[0020] In one embodiment, the process of obtaining each sample time change vector and each sample feature change vector is as follows:

[0021] The coordinates composed of the attribute information of each sanitation equipment at each inspection are recorded as feature coordinates; the set composed of all feature coordinates of all sanitation equipment of the same type is recorded as a feature data set; the vector composed of the maintenance interval time obtained when any two feature coordinate points in the feature data set correspond to the inspection is recorded as a sample time change vector;

[0022] Between any two characteristic coordinate points, a vector formed by taking the characteristic coordinate point corresponding to the minimum maintenance interval time as the starting point and taking the characteristic coordinate point corresponding to the maximum maintenance interval time as the end point is recorded as a sample characteristic change vector.

[0023] In one embodiment, the process of predicting the feature change vector of any pending device is:

[0024] The time spent on the journey from the sanitation equipment being inspected at the current moment to any pending equipment is recorded as the inspection time of any pending equipment;

[0025] Calculate the sum of the real-time maintenance interval time and the inspection time of any pending device at the current moment, record it as a first sum, and use the vector composed of the real-time maintenance interval time of any pending device at the current moment and the first sum as the real-time time change vector of any pending device;

[0026] Calculate the similarity between the real-time time change vector of any pending device at the current moment and each sample time change vector in the feature data set where the any pending device is located, recorded as a first similarity; use the normalized value of the first similarity as the time similarity weight of the sample feature change vector corresponding to each sample time change vector;

[0027] Calculate the product of each sample feature change vector in the feature data set where any of the pending devices is located and the corresponding time similarity weight, and record it as the second product; and take the sum of all the second products of any of the pending devices as the feature change vector of any of the pending devices predicted at the current moment.

[0028] In one embodiment, the process of acquiring the predicted detection vector of any pending device is as follows:

[0029] The first sum value of any pending device is used as the predicted maintenance interval of any pending device; the sum of the real-time inspection interval of any pending device at the current moment and the inspection time is used as the predicted inspection interval of any pending device; the real-time attribute information of any pending device at the current moment is summed with the corresponding elements in the feature change vector to obtain the predicted attribute information of any pending device;

[0030] A vector consisting of the predicted attribute information, predicted inspection interval time and predicted maintenance interval time of any pending device is recorded as a predicted detection vector of any pending device.

[0031] In one embodiment, the process of obtaining the inspection scheme evaluation value is as follows:

[0032] The sanitation equipment inspected at the current moment is recorded as , then record the current state as state ; Calculation status The sum of the device abnormality values ​​of all pending devices in the park is used as the status The state of abnormal value;

[0033] For any inspection plan, the first pending device to be inspected in the inspection plan is recorded as , the second pending device to be inspected is recorded as ;from Inspection to The time is recorded as At this moment, the corresponding state is recorded as state ;from Inspection to The time is recorded as At this moment, the corresponding state is recorded as state ;

[0034] The status The predicted abnormal value of each pending device obtained at the time is used as the state Device abnormal value of each pending device; status When calculating the device and The sum of the device abnormality values ​​of all other pending devices in the park other than The state of abnormal value;

[0035] The status When each device is The predicted detection vector at time is used as the state Real-time detection vector of each device, calculation status When each device is The predicted anomaly value at time t, as the state Device abnormal value of each device at the time; status When calculating the device , and The sum of the abnormal values ​​of all other devices in the park except The state of abnormal value;

[0036] Similarly, the abnormal status values ​​corresponding to each pending device in the inspection plan vector are obtained in turn; The state abnormality values ​​corresponding to all pending devices in the inspection plan vector form a state abnormality value sequence, calculate the difference between each element in the state abnormality value sequence and the next element, and record the ratio of the difference to the corresponding inspection time as the abnormal value decrease rate; the sum of the decrease rates of all abnormal values ​​in the state abnormality value sequence is used as the inspection plan evaluation value of any inspection plan.

[0037] In one embodiment, the process of obtaining the best inspection plan for inspecting sanitation equipment is as follows:

[0038] The inspection scheme evaluation value of each inspection scheme is used as the adaptability, and a genetic algorithm is used to screen all inspection schemes to obtain the best inspection scheme; the first device in the best inspection scheme is inspected.

[0039] This application has at least the following beneficial effects:

[0040] This application calculates the equipment abnormality values ​​of different sanitation equipment through historical inspection data to characterize the urgency of the need for inspection of sanitation equipment; calculates the predicted abnormality values ​​through the change characteristics of the detection data of different sanitation equipment, combined with the inspection time, to characterize the estimated value of the equipment abnormality value of the sanitation equipment after a certain delay, so as to avoid the inspection plan evaluation error caused by ignoring the time spent by the inspection personnel between the sanitation equipment when judging the effects of different inspection plans in the future; constructs an inspection plan vector, calculates the inspection plan evaluation value according to the impact of the inspection plan on the equipment abnormality value of the entire inspection area, and selects the best inspection plan in combination with the genetic algorithm; wherein, by constructing the inspection plan vector, the equipment abnormality values ​​of the sanitation equipment around the next inspection target are evaluated to avoid the situation where the equipment abnormality value of the selected target path value itself is high, but the surrounding equipment abnormality values ​​are low, resulting in the inspection plan not conforming to the overall interests; finally, according to the selected best inspection plan, guides the inspection personnel to complete the inspection. Among them, since this application makes full use of the intelligent information of the equipment when obtaining the inspection plan, and uses the equipment abnormal values ​​to characterize the urgency of inspection of different sanitation equipment, and plans the inspection path of the sanitation equipment in combination with the distance between different sanitation equipment, compared with the traditional fixed inspection plan, it can dynamically perform equipment inspections according to the situation of the sanitation equipment, avoiding the waste of inspection resources and the occurrence of inspection blind spots due to improper allocation of equipment inspection resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0042] Figure 1 A flow chart of the intelligent inspection method for park scenic spots based on environmental sanitation integration provided in this application;

[0043] Figure 2 Schematic diagram of the process of predicting the characteristic change vector of any pending device. DETAILED DESCRIPTION

[0044] In order to further explain the technical means and effects adopted by this application to achieve the predetermined invention purpose, the specific implementation method, structure, characteristics and effects of the intelligent inspection method for park scenic spots based on sanitation integration proposed in this application are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0045] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0046] The specific scheme of the intelligent inspection method for park scenic spots based on integrated sanitation provided by this application is described in detail below with reference to the accompanying drawings.

[0047] An embodiment of the present application provides an intelligent inspection method for park scenic spots based on integrated sanitation.

[0048] Specifically, the following park scenic area intelligent inspection method based on environmental sanitation integration is provided, please refer to Figure 1 , the method comprises the following steps:

[0049] Step S1, collecting various types of inspection information of each type of sanitation equipment, including attribute information and time information of the sanitation equipment.

[0050] (1) The sanitation facilities that need to be inspected in the embodiment of the present application include garbage collection facilities, public health facilities, garbage transfer facilities and irrigation facilities. The attribute information of each sanitation equipment of each sanitation facility during the inspection is obtained, specifically:

[0051] The garbage collection facilities include various smart garbage bins. The weight of garbage in each smart garbage bin is detected by the pressure sensor of each smart garbage bin. Since each place where garbage bins are set usually contains two garbage bins, recyclable and non-recyclable, each smart garbage bin can obtain two garbage weight data at any time, corresponding to the weight of recyclable garbage and the weight of non-recyclable garbage respectively;

[0052] Public sanitation facilities include public toilets, and the air humidity and ammonia concentration of each public toilet are detected by air humidity sensors and air quality sensors in each public toilet respectively;

[0053] The garbage transfer facilities include various garbage transfer stations. The pressure sensor and infrared sensor of each garbage transfer station are used to detect the garbage weight and garbage height in each garbage transfer station, obtain the depth of the garbage bin of the garbage transfer station, and calculate the ratio of the garbage height to the garbage bin depth of the garbage transfer station to obtain the volume capacity ratio of each garbage transfer station.

[0054] The irrigation facility includes various irrigation devices, and the temperature and humidity of the soil are detected respectively by a temperature sensor and a humidity sensor in each irrigation device.

[0055] (2) The two types of data collected by each sanitation equipment of each sanitation facility are used as the attribute information of each sanitation equipment of each sanitation facility, and the attribute information of the sanitation equipment collected during each inspection is stored in the smart sanitation system for easy data retrieval. Among them, since each inspection can only check one sanitation equipment, the data collected during each inspection in this application is only the attribute information data of one sanitation equipment.

[0056] (3) Obtain the time information of the sanitation equipment inspected during each inspection: obtain the time difference between the current inspection time of the sanitation equipment and the time of its last maintenance, and record it as the maintenance interval time of the sanitation equipment at the time of the inspection; obtain the time difference between the current inspection time of the sanitation equipment and the time of its last inspection, and record it as the inspection interval time of the sanitation equipment at the time of the inspection. Similarly, the time information of the sanitation equipment collected during each inspection is stored in the smart sanitation system for data retrieval.

[0057] The attribute information and time information of each sanitation equipment are used as the inspection information of each sanitation equipment.

[0058] (4) Furthermore, the historical inspection information of each sanitation equipment of each sanitation facility is obtained through the smart sanitation system, specifically:

[0059] First, the maintenance status of the sanitation equipment during each inspection is obtained. If maintenance is performed, the tag value of the sanitation equipment during this inspection is set to 1, otherwise it is set to 0.

[0060] Then, for each sanitation equipment, in the inspection information of the sanitation equipment recorded by the smart sanitation system, obtain the Q inspection information with the label value of 1 closest to the current moment, and the Q inspection information with the label value of 0 closest to the current moment, and use these 2Q inspection information as the historical inspection information of the sanitation equipment. Preferably, in the embodiment of the present application, the value of Q is set to 20. As other embodiments of the present application, the implementer can set the value of Q according to the actual situation.

[0061] Step S2, construct each historical detection vector of each sanitation equipment based on the inspection information of each sanitation equipment each time it is inspected in historical time; obtain the inspection information of each sanitation equipment at the current moment to construct the real-time detection vector of each sanitation equipment; record the remaining sanitation equipment except the sanitation equipment inspected at the current moment as pending equipment; based on the difference between the real-time detection vector of each pending equipment and the historical detection vector of the same sanitation equipment, as well as the maintenance status of the same sanitation equipment each time it is inspected, construct the equipment abnormal value of each pending equipment.

[0062] During inspection, after the current sanitation equipment has been inspected, the next sanitation equipment needs to be inspected. To determine the next sanitation equipment to be inspected, first obtain the historical inspection data and the most recent inspection data of each sanitation equipment, and calculate the abnormal value of each sanitation equipment, specifically:

[0063] (1) In the historical inspection information of each sanitation equipment, the vector composed of the inspection information collected each time the sanitation equipment is inspected is used as the historical detection vector of the current inspection of the sanitation equipment. Taking public toilet A as an example, if the sanitation equipment inspected this time is public toilet A, then the historical detection vector of the current inspection of the sanitation equipment is the vector composed of the air humidity, ammonia concentration, maintenance interval and inspection interval of public toilet A during this inspection.

[0064] (2) At the current moment, the data of the sensors of each sanitation equipment is obtained to obtain the real-time attribute information of each sanitation equipment at the current moment. Taking public toilet A as an example, the real-time attribute information of public toilet A is the air humidity and ammonia concentration data detected by the sensors in public toilet A at the current moment.

[0065] At the same time, the time interval between the current moment and the time when each sanitation equipment was last inspected is calculated as the real-time inspection interval of each sanitation equipment at the current moment; and the time interval between the current moment and the time when each sanitation equipment was last maintained is calculated as the real-time maintenance interval of each sanitation equipment at the current moment.

[0066] The vector composed of the real-time attribute information, real-time inspection interval and real-time maintenance interval of each sanitation equipment is used as the real-time detection vector of each sanitation equipment at the current moment.

[0067] (3) Calculate equipment abnormal values:

[0068] In sanitation inspection work, fixed inspection plans adopt fixed route inspections, ignoring the monitoring information of intelligent sanitation equipment, resulting in the lack of intercommunication between offline inspection and online monitoring data, and further leading to unreasonable allocation of inspection resources.

[0069] (3.1) At the current moment, the inspection personnel need to move from the sanitation equipment that has just been inspected to other sanitation equipment for inspection. In order to determine the next sanitation equipment to be inspected, each sanitation equipment except the sanitation equipment that has just been inspected at the current moment is recorded as a pending equipment.

[0070] (3.2) The device abnormality value of each pending device is calculated based on the difference between the real-time detection vector of each pending device and all historical detection vectors of all sanitation devices of the same type, so as to evaluate the inspection priority of different sanitation devices. The calculation method is:

[0071] First, the Euclidean distance between the real-time detection vector of each pending device and each historical detection vector of each sanitation device of the same type is calculated, recorded as the first Euclidean distance, and normalized.

[0072] Preferably, in the embodiment of the present application, the normalization method is: calculating the sum of all the first Euclidean distances of each device to be determined, and taking the ratio of each first Euclidean distance of each device to be determined to the sum as the normalized value of each first Euclidean distance of each device to be determined. It should be noted that there are many existing normalization methods, and the implementer may also use other normalization algorithms to normalize the first Euclidean distance, and the present application does not make specific restrictions.

[0073] Then, the normalized value of each of the first Euclidean distances of each pending device is used as the distance weight of the corresponding historical detection vector, and the product of the label value corresponding to each historical detection vector of each pending device and the corresponding distance weight is recorded as the first product; the sum of the first products of all historical detection vectors of each pending device is used as the device anomaly value of each pending device.

[0074] The larger the abnormal value of each pending device is, the more likely it is that the pending device is a sanitation device that needs maintenance at this time, and the pending device should be inspected first during the inspection.

[0075] Step S3, based on the change amount between the same type of elements in the historical detection vectors of any two sanitation equipment of the same type, construct each sample time change vector and each sample feature change vector of each type of sanitation equipment, combine the time required for the sanitation equipment inspected at the current moment to reach any pending equipment, and the time information of any pending equipment at the current moment, predict the feature change vector of any pending equipment; construct the predicted detection vector of any pending equipment based on the real-time detection vector and the predicted feature change vector of any pending equipment, and obtain the predicted abnormal value of any pending equipment by calculating the equipment abnormal value.

[0076] The maintenance intervals of different sanitation equipment are different. Some sanitation equipment requires frequent maintenance, while some sanitation equipment can continue to work for a long time without maintenance. For a sanitation equipment, its abnormal value may be small at the current moment, but its abnormal value increases very quickly. Therefore, when conducting inspections, it is necessary to give priority to sanitation equipment with a faster increase in abnormal values.

[0077] Especially when selecting the inspection route, it will take a certain amount of time for the inspectors to travel from the current sanitation equipment to the next one. The increase in abnormal values ​​during this period needs to be considered, otherwise it will cause errors in the estimation of abnormal values, which will lead to inspection blind spots where some equipment cannot be inspected in time.

[0078] Therefore, (1) the coordinates consisting of the two types of attribute information data collected each time each sanitation equipment is inspected are marked as feature coordinates. Taking public toilet A as an example, each feature coordinate of public toilet A is a coordinate consisting of the air humidity and ammonia concentration of public toilet A each time it is inspected. The set of all feature coordinates of all sanitation equipment of each type of sanitation facility is recorded as the feature data set of this type of sanitation facility.

[0079] (2) For the characteristic data sets of various sanitation facilities, the maintenance intervals corresponding to any two characteristic coordinate points in the characteristic data sets are arranged in order from small to large, and the resulting vector is recorded as the sample time change vector , and at the same time The corresponding feature coordinate point is taken as the starting point and The vector formed by the corresponding feature coordinate point as the end point is recorded as the sample feature change vector Assume that the number of sample time change vectors and the number of sample feature change vectors in the feature data sets of various sanitation facilities are both Q.

[0080] (3) Record the current time as time t, and obtain the time it takes to travel from the sanitation equipment just inspected to the nth undetermined equipment, which is recorded as the inspection time ;

[0081] Furthermore, the real-time maintenance interval of the nth pending equipment at the current moment is recorded as , then in At time , the real-time maintenance interval of the nth pending equipment is , further, the real-time time change vector of the nth pending device at the current moment is recorded as ;

[0082] (4) Using the above parameters, calculate the current time nth pending device The characteristic change vector at the moment represents the prediction of the nth pending device at the current moment after the inspection time through historical data. After that, the change of the attribute information data of the nth pending device is:

[0083] First, the nth pending device at the current moment is Real-time time change vector at the moment The cosine similarity between the time change vector of each sample in the feature data set of the nth undetermined device is recorded as the first similarity. The larger the cosine similarity of each sample time change vector, the more the corresponding time change conforms to the time change of the nth undetermined device from time t to The more time changes at each moment, the more we can use the corresponding sample feature change vector to predict The real-time detection vector of the nth pending device at time.

[0084] It should be noted that for the similarity calculation between the real-time time change vector and the sample time change vector, this application only provides a similarity calculation method. There are many existing similarity calculation methods. Implementers can also use other similarity algorithms to obtain the similarity between the real-time time change vector and the sample time change vector. This application does not make specific restrictions.

[0085] Then, all the first similarities of the nth device to be determined are non-negative and normalized by adding 1 to each of the first similarities to complete the non-negative process; further calculating the sum of all the non-negative first similarities, and normalizing each non-negative first similarity by the sum to obtain a value as the time similarity weight of the sample feature change vector corresponding to each sample time change vector;

[0086] Finally, the product of each sample feature change vector in the feature data set where the nth pending device is located and its corresponding time similarity weight is calculated, which is recorded as the second product; the sum of all the second products of the nth pending device is used as the predicted nth pending device at the current moment. The feature change vector at the moment.

[0087] (5) The feature change vector is the real-time detection vector predicted by historical sample data. The eigenvalue at the moment, so through the characteristic change vector Can The real-time detection vector of the nth pending device at the moment is predicted by:

[0088] Get the real-time maintenance interval plus inspection time of the nth pending equipment at the current moment Get the predicted maintenance interval of the nth pending equipment; add the real-time inspection interval of the nth pending equipment at the current moment to the inspection time Obtain the predicted inspection interval time of the nth pending device; sum the two real-time attribute information of the nth pending device at the current moment with the same elements in the characteristic change vector to obtain the two predicted attribute information of the nth pending device;

[0089] Furthermore, the vector consisting of the two types of predicted attribute information, the predicted inspection interval time and the predicted maintenance interval time of the nth pending equipment is recorded as the predicted detection vector of the nth pending equipment .

[0090] Furthermore, the feature change vector is the data change feature extracted from the historical data of the inspection data set, which represents the time from time t to The characteristic change law of the nth pending device at the moment, and the resulting predicted detection vector can be used as the evaluation Therefore, based on the predicted detection vector of the nth pending device and all its historical detection vectors, the same calculation method as the device abnormality value of each pending device is used to obtain the device abnormality value of the nth pending device at time Predicted outliers at time , so as to predict the nth pending device at the current moment The abnormal value of the device at the moment.

[0091] Step S4, randomly generate a preset number of inspection plans, calculate the inspection plan evaluation value of each inspection plan based on the equipment abnormal value and predicted abnormal value of the pending equipment, and obtain the best inspection plan to inspect the sanitation equipment.

[0092] (1) At the current moment, the inspection personnel need to inspect the pending devices. First, they number all the pending devices. Then, they randomly generate an inspection plan with a length of A = 5, and record the vector composed of all the pending devices in the inspection plan in the inspection order as the inspection plan vector ,in, , , , and Respectively represent , , , First pending equipment; record the sanitation equipment that has just been inspected at the current moment as , then the inspection plan vector at the current moment is The inspection path represented by It should be noted that, for the value of the length A, the implementer can set the value of A according to the actual situation, and this application does not impose any specific restrictions.

[0093] (2) For the randomly generated inspection plan vector , calculate the rate of decline of its abnormal value as the standard for evaluating the quality of the inspection plan, the calculation method is:

[0094] First, get sanitation equipment To pending device The inspection time is recorded as ;

[0095] Then, the simulated patrol personnel choose the path After the inspection, the abnormal value changes in the entire inspection area are calculated as follows:

[0096] At this moment, the inspection personnel are At this point, record the current state as state , obtain the real-time detection vector of each pending device, which is used to obtain the device abnormality value of each pending device; calculate the sum of the device abnormality values ​​of all pending devices in the park as the state The state of abnormal value;

[0097] Further, in At all times, the inspectors are on the equipment , and record the state at this time as state , the state The predicted abnormal value of each pending device is used as the state Device abnormal value of each pending device; status When calculating the device and equipment The sum of the device abnormality values ​​of all other pending devices in the park other than The state of abnormal value;

[0098] Further, get the status After the state outlier value of Status of abnormal values, methods and status The method for obtaining the state abnormal value is similar, specifically:

[0099] Get the device To device The inspection time is recorded as ;

[0100] The status When each device is The predicted detection vector at time is used as the state Real-time detection vector of each device; according to the status The real-time detection vector of each device and all the historical detection vectors corresponding to each device are used in conjunction with the state The same method is used to obtain the predicted detection vectors of each pending device, and the state is calculated. When each device is The predicted detection vector at time t, as the state Real-time detection vector of each device;

[0101] According to the status The real-time detection vector of each device and all the historical detection vectors corresponding to each device are used in conjunction with the state The same calculation method is used for the abnormal value of each pending device, and the calculation status is Device abnormal value of each device at the time; status When calculating the device ,equipment and equipment The sum of the abnormal values ​​of all other devices in the park except The state of abnormal value;

[0102] And so on, get the status in sequence , and The status of abnormal value.

[0103] Finally, the inspection plan vector , stateful , , , , , There are A+1 state anomaly values ​​in total. Arrange the above A+1 state anomaly values ​​in the inspection order to form a state anomaly value sequence. Then, the adjacent state anomaly values ​​in the sequence are subtracted from the left element by the right element to obtain the difference, and the difference is divided by the corresponding inspection time to obtain A ratios, where the uth ratio represents the rate of decrease of the anomaly value of the uth path;

[0104] Among them, the difference in state anomaly values ​​represents the change in anomaly values ​​of the entire inspection area before and after selecting a certain path. The more the anomaly value decreases, the more conducive it is to reduce the anomaly value of the entire area by selecting this path, and selecting this path is the best path selection result. The difference is compared with the inspection time because the inspection time is the price paid for selecting a path. The smaller the price paid and the more the anomaly value decreases, the faster the anomaly value decreases, and the selected path is a better path solution.

[0105] (3) Sum the decreasing speeds of A outliers to obtain the inspection plan vector The larger the inspection plan evaluation value, the more the inspection plan can reduce the overall abnormal value of all equipment.

[0106] It is particularly noted that, in the inspection plan vector generated at the current moment in this embodiment, only the first element will actually guide the selection of the inspection target, and the remaining elements may not necessarily become the actual path selection plan; for example, if the inspection plan vector is the selected inspection plan, then It is the next inspection target. After that, it will reselect the next inspection target, not necessarily As an inspection target.

[0107] When performing inspection according to the inspection plan vector, after the inspection of the equipment After that, the entire inspection area is in a state ,At this time, a new inspection path is regenerated and optimized to obtain the next inspection target.

[0108] For the inspection plan vector The non-first element in is calculated to evaluate: As the next inspection target, As the starting point for the next path selection, The selection of surrounding paths. For example, the current inspection personnel are on the sanitation equipment There are two sanitation equipments with opposite directions to choose from. , , where the path The corresponding outlier value decreases faster than the path The corresponding outlier value decreases at a rate of The outliers of the neighboring sanitation equipment of the first sanitation equipment are generally higher, while the The outliers of the neighboring sanitation equipment are generally low. , then the subsequent path selection situation cannot quickly reduce the overall abnormal value of the entire inspection area; on the contrary, the path selection ,Although the rate of decreasing of outliers in this selection is not the fastest, the subsequent ,selection of adjacent sanitation equipment can rapidly reduce the overall outliers in the ,entire inspection area.

[0109] (4) This embodiment uses the inspection scheme evaluation value as the adaptability and adopts a genetic algorithm to screen the randomly generated inspection schemes, specifically:

[0110] P1 initialization: First, randomly generate C = 50 inspection plan vectors. The generation conditions of the inspection plan vectors are: the vector length is A, the vector elements are integers from 1 to S, and the vectors cannot be repeated. Then the set of C inspection plan vectors is used as a plan data set.

[0111] P2 screening: Calculate the inspection plan evaluation value for all inspection plan vectors in the plan data set, select D=5 inspection plan vectors with the smallest inspection plan evaluation value as the winning individuals, and take the set of D winning individuals as a winning individual set;

[0112] P3 crossover mutation: crossover and mutation are performed on the winning individuals, with a crossover rate of 0.10, a mutation rate of 0.02, and C = 50 crossover mutations. The crossover mutation process is a well-known technology in the field of data processing. Finally, C inspection plan vectors are output, and the set of C inspection plan vectors is regarded as a plan data set;

[0113] P4 loop: If the number of loops is less than E=50 times, obtain the solution data set obtained in step P3 and repeat steps P2 and P3. Otherwise, select an inspection solution vector with the largest inspection solution evaluation value from the solution data set as the best inspection solution.

[0114] Finally, the first element of the best inspection plan is obtained as the next sanitation equipment to be inspected. Among them, steps P1 to P4 are well-known contents. It should be noted that in other embodiments of the present application, the implementer can set the value of each parameter in steps P1 to P4 according to the actual situation, and the present application does not make specific restrictions.

[0115] In the actual operation process, when the inspection personnel complete the sanitation equipment After the inspection, a request is sent to the smart sanitation system, instructing the sanitation system to obtain the sanitation equipment for the next inspection of the inspector according to this embodiment, and complete the instructions to the inspector.

[0116] The schematic diagram of the process of predicting the characteristic change vector of any pending device is as follows: Figure 2 shown.

[0117] To summarize, the embodiment of the present application calculates the equipment abnormality values ​​of different sanitation equipment through historical inspection data to characterize the urgency of the need for inspection of the sanitation equipment; calculates the predicted abnormality values ​​through the change characteristics of the detection data of different sanitation equipment in combination with the inspection time to characterize the estimated value of the equipment abnormality value of the sanitation equipment after a certain delay, so as to avoid the inspection plan evaluation error caused by ignoring the time spent by the inspection personnel between the sanitation equipment when judging the effects of different inspection plans in the subsequent process; constructs an inspection plan vector, calculates the inspection plan evaluation value according to the impact of the inspection plan on the equipment abnormality value of the entire inspection area, and selects the best inspection plan in combination with the genetic algorithm; wherein, by constructing the inspection plan vector, the equipment abnormality values ​​of the sanitation equipment around the next inspection target are evaluated to avoid the situation where the equipment abnormality value of the selected target path value itself is high, but the surrounding equipment abnormality values ​​are low, resulting in the inspection plan not conforming to the overall interests; finally, according to the selected best inspection plan, guides the inspection personnel to complete the inspection. Among them, since this application makes full use of the intelligent information of the equipment when obtaining the inspection plan, the urgency of inspection of different sanitation equipment is represented by the equipment abnormal values, and the inspection path of the sanitation equipment is planned in combination with the distance between different sanitation equipment. Compared with the traditional fixed inspection plan, it can dynamically inspect the equipment according to the situation of the sanitation equipment, avoiding the waste of inspection resources and the occurrence of inspection blind spots due to improper allocation of equipment inspection resources.

[0118] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. The above-mentioned specific embodiments of the present application are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0119] The various embodiments in the present application are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0120] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Modifications to the technical solutions recorded in the aforementioned embodiments, or equivalent replacement of some of the technical features therein, do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. An intelligent inspection method for park scenic spots based on environmental sanitation integration is characterized by: The method comprises the following steps: Collect various inspection information of each type of sanitation equipment, including attribute information and time information of sanitation equipment; Construct each historical detection vector of each sanitation equipment based on the inspection information of each sanitation equipment at each inspection in the historical time; obtain the inspection information of each sanitation equipment at the current moment to construct the real-time detection vector of each sanitation equipment; record the remaining sanitation equipment except the sanitation equipment inspected at the current moment as pending equipment; construct the equipment abnormal value of each pending equipment based on the difference between the real-time detection vector of each pending equipment and the historical detection vector of the same sanitation equipment, as well as the maintenance status of the same sanitation equipment at each inspection; Based on the variation between the same type of elements in the historical detection vectors of any two sanitation equipment of the same type, each sample time variation vector and each sample feature variation vector of each type of sanitation equipment are constructed. The feature variation vector of any pending equipment is predicted by combining the time required for the sanitation equipment inspected at the current moment to reach any pending equipment and the time information of any pending equipment at the current moment. Based on the real-time detection vector and the predicted feature variation vector of any pending equipment, a predicted detection vector of any pending equipment is constructed, and the predicted abnormal value of any pending equipment is obtained by calculating the abnormal value of the equipment. Randomly generate a preset number of inspection plans, calculate the inspection plan evaluation value of each inspection plan based on the equipment abnormal value and predicted abnormal value of the pending equipment, and obtain the best inspection plan for sanitation equipment inspection; The process of obtaining the device abnormal value of each pending device is as follows: If the sanitation equipment is repaired each time it is inspected, the tag value of the sanitation equipment is set to 1 each time it is inspected, otherwise it is set to 0; Calculate the Euclidean distance between the real-time detection vector of each pending device and each historical detection vector of each sanitation device of the same type, and record it as the first Euclidean distance; use the normalized value of each of the first Euclidean distances of each pending device as the distance weight of the corresponding historical detection vector, and record the product of the label value corresponding to each historical detection vector of each pending device and the corresponding distance weight as the first product; and use the sum of the first products of all historical detection vectors of each pending device as the device abnormality value of each pending device.

2. The intelligent inspection method for park scenic spots based on environmental sanitation integration as claimed in claim 1 is characterized in that: The time information is: The time difference between each inspection time of each sanitation equipment and the corresponding last maintenance time is recorded as the maintenance interval time of each sanitation equipment when it is inspected; And the time difference between each inspection time of each sanitation equipment and the corresponding last inspection time.

3. The intelligent inspection method for park scenic spots based on environmental sanitation integration as claimed in claim 1 is characterized in that: Each historical detection vector of each sanitation equipment is a vector composed of all attribute information and all time information of each sanitation equipment each time it is inspected.

4. The intelligent inspection method for park scenic spots based on environmental sanitation integration as claimed in claim 1 is characterized in that: The acquisition process of the real-time detection vector of each sanitation equipment is as follows: The attribute information of each sanitation equipment at the current moment is used as the real-time attribute information of each sanitation equipment; the time interval between the current moment and the time when each sanitation equipment was last inspected is used as the real-time inspection interval of each sanitation equipment; the time interval between the current moment and the time when each sanitation equipment was last repaired is used as the real-time maintenance interval of each sanitation equipment; The vector composed of the real-time attribute information, real-time inspection interval and real-time maintenance interval of each sanitation equipment is used as the real-time detection vector of each sanitation equipment.

5. The intelligent inspection method for park scenic spots based on environmental sanitation integration as claimed in claim 1 is characterized in that: The acquisition process of each sample time change vector and each sample feature change vector is as follows: The coordinates composed of the attribute information of each sanitation equipment at each inspection are recorded as feature coordinates; the set composed of all feature coordinates of all sanitation equipment of the same type is recorded as a feature data set; the vector composed of the maintenance interval time obtained when any two feature coordinate points in the feature data set correspond to the inspection is recorded as a sample time change vector; Between any two characteristic coordinate points, a vector formed by taking the characteristic coordinate point corresponding to the minimum maintenance interval time as the starting point and taking the characteristic coordinate point corresponding to the maximum maintenance interval time as the end point is recorded as a sample characteristic change vector.

6. The intelligent inspection method for park scenic spots based on environmental sanitation integration as claimed in claim 4 is characterized in that: The process of predicting the characteristic change vector of any pending device is as follows: The time spent on the journey from the sanitation equipment being inspected at the current moment to any pending equipment is recorded as the inspection time of any pending equipment; Calculate the sum of the real-time maintenance interval time and the inspection time of any pending device at the current moment, record it as a first sum, and use the vector composed of the real-time maintenance interval time of any pending device at the current moment and the first sum as the real-time time change vector of any pending device; Calculate the similarity between the real-time time change vector of any pending device at the current moment and each sample time change vector in the feature data set where the any pending device is located, recorded as a first similarity; use the normalized value of the first similarity as the time similarity weight of the sample feature change vector corresponding to each sample time change vector; Calculate the product of each sample feature change vector in the feature data set where any of the pending devices is located and the corresponding time similarity weight, and record it as the second product; and take the sum of all the second products of any of the pending devices as the feature change vector of any of the pending devices predicted at the current moment.

7. The intelligent inspection method for park scenic spots based on environmental sanitation integration as claimed in claim 6 is characterized in that: The process of obtaining the predicted detection vector of any pending device is as follows: Using the first sum value of any pending equipment as the predicted maintenance interval time of any pending equipment; The sum of the real-time inspection interval of any pending device at the current moment and the inspection time is used as the predicted inspection interval of any pending device; The real-time attribute information of any pending device at the current moment is summed with the corresponding elements in the feature change vector to obtain the predicted attribute information of any pending device; A vector consisting of the predicted attribute information, predicted inspection interval time and predicted maintenance interval time of any pending device is recorded as a predicted detection vector of any pending device.

8. The intelligent inspection method for park scenic spots based on environmental sanitation integration as claimed in claim 6 is characterized in that: The process of obtaining the inspection scheme evaluation value is as follows: The sanitation equipment inspected at the current moment is recorded as , then record the current state as state ; Calculation status The sum of the device abnormality values ​​of all pending devices in the park is used as the status The state of abnormal value; For any inspection plan, the first pending device to be inspected in the inspection plan is recorded as , the second pending device to be inspected is recorded as ;from Inspection to The time is recorded as At this moment, the corresponding state is recorded as state ;from Inspection to The time is recorded as At this moment, the corresponding state is recorded as state ; The status The predicted abnormal value of each pending device obtained at the time is used as the state Device abnormal value of each pending device; status When calculating the device and The sum of the device abnormality values ​​of all other pending devices in the park other than The state of abnormal value; The status When each device is The predicted detection vector at the moment is used as the state Real-time detection vector of each device, calculation status When each device is The predicted anomaly value at time t, as the state Device abnormal value of each device at the time; status When calculating the device , and The sum of the abnormal values ​​of all other devices in the park except The state of abnormal value; Similarly, the abnormal status values ​​corresponding to each pending device in the inspection plan vector are obtained in turn; The state abnormality values ​​corresponding to all pending devices in the inspection plan vector form a state abnormality value sequence, calculate the difference between each element in the state abnormality value sequence and the next element, and record the ratio of the difference to the corresponding inspection time as the abnormal value decrease rate; the sum of the decrease rates of all abnormal values ​​in the state abnormality value sequence is used as the inspection plan evaluation value of any inspection plan.

9. The intelligent inspection method for park scenic spots based on environmental sanitation integration as claimed in claim 8 is characterized in that: The process of obtaining the best inspection plan for sanitation equipment inspection is as follows: The inspection scheme evaluation value of each inspection scheme is used as the adaptability, and a genetic algorithm is used to screen all inspection schemes to obtain the best inspection scheme; the first device in the best inspection scheme is inspected.

Images