Water quality safety multi-parameter linkage early warning system based on fish behavior
By comprehensively processing multiple parameters in fish behavior, a multi-parameter linkage early warning system based on fish behavior was designed, which solves the problem of false alarms caused by interference factors in the existing technology and improves the accuracy and safety of water quality monitoring.
Patent Information
- Application Number
- CN202211561654.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-12-07
- Publication Date
- 2025-12-12
- Estimated Expiration
- 2042-12-07
AI Technical Summary
Existing technologies that use fish behavior to reflect water quality are easily affected by interference factors such as light, sound, and vibration, leading to false alarms or low accuracy, and cannot effectively consider the intrinsic relationship between multiple parameters.
Design a multi-parameter linkage early warning system based on fish behavior, including a water tank, a camera, an initial speed early warning module, a final speed early warning module, an avoidance rate early warning module, a two-parameter linkage early warning module, and a multi-parameter linkage early warning module. By comprehensively processing parameters such as the average speed and avoidance rate of fish, the system generates the final water quality judgment result.
It improves the accuracy of water quality monitoring, reduces the false alarm rate, and can effectively monitor water quality safety and protect human life and property.
Smart Images

Figure CN116453295B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the field of water quality monitoring, in particular to a water quality safety multi-parameter linkage early warning system based on fish ethology. BACKGROUND
[0002] The present application makes full use of the close relationship between the behavior of aquatic organisms (such as fish) and the water quality environment, and through real-time monitoring and analysis of fish school behavior by computer vision, the water quality of the water body can be reflected and early warning can be made.
[0003] In most cases, such fish ethology parameters directly reflect the water quality condition, and the existing technology generally selects a single parameter or a simple multi-parameter such as speed, acceleration, fish school spacing, and avoidance rate as the basis for reflecting the water quality condition, and obtains the corresponding threshold data under different toxicant concentrations through experiments.
[0004] However, fish biological behavior is inevitably affected by various interference factors (such as light, sound, vibration, temperature, etc.), and traditional early warning technology is prone to false alarms or low accuracy. Therefore, the mutual relationship of multi-parameters within a certain time span and space span should be fully considered.
[0005] For example, a time span behavior is that when the toxicant dose is large, the fish school will produce a significant stress response rapidly and cause a significant change in the corresponding behavior parameter, and then transition to slow change; when the toxicant dose is small, the fish school will cause a sustained slow change in the corresponding behavior parameter after long-term exposure to the toxicant. For another example, a space span behavior is that in a monitoring system implementing an avoidance zone (a safe zone containing clean water flow is arranged therein), the fish school behavior has no difference in the entire water body space under the non-toxic state, but there is a significant behavior difference inside and outside the avoidance zone under the toxic state.
[0006] Therefore, a multi-parameter linkage early warning system considering the internal relationship between parameters is needed. SUMMARY
[0007] The purpose of the present application is to provide a water quality safety multi-parameter linkage early warning system based on fish ethology, which comprises a water tank, a camera, an initial speed early warning module, a final speed early warning module, an avoidance rate early warning module, a two-parameter linkage early warning module, and a multi-parameter linkage early warning module.
[0008] The water tank is used for group feeding of fish.
[0009] The water tank comprises a fish school activity area and an avoidance zone; the water in the fish school activity area is a detection water sample; and the water in the avoidance zone is a normal water sample.
[0010] The camera shoots fish activity images and transmits them to the initial speed early warning module, the final speed early warning module, and the avoidance rate early warning module, respectively.
[0011] The initial speed warning module stores an initial speed judgment index, including an initial speed abnormal threshold TV1, an initial speed danger threshold TV2, an initial speed abnormal time threshold T1, and an initial speed danger time threshold T2.
[0012] The initial speed warning module processes the fish activity image to obtain the average speed V of the fish and transmits it to the final speed warning module.
[0013] The initial speed warning module determines the relationship between the average speed V of the fish and the initial speed judgment index to obtain the water quality judgment result S1 and transmits it to the two-parameter linkage warning module.
[0014] The avoidance rate warning module stores an avoidance rate detection period T4, an avoidance rate determination period T5, and an avoidance rate judgment index.
[0015] The avoidance rate judgment index includes an avoidance rate abnormal threshold TR1, a danger threshold TR2, an abnormal number threshold TRN1 within the avoidance rate determination period T5, and a danger number threshold TRN2 within the avoidance rate determination period T5.
[0016] The avoidance rate warning module processes the fish activity image according to the avoidance rate detection period T4 to obtain the fish avoidance rate R.
[0017] The avoidance rate warning module determines the relationship between the fish avoidance rate R and the avoidance rate judgment index to obtain the water quality judgment result S2 and transmits it to the two-parameter linkage warning module.
[0018] The final speed warning module stores a final speed judgment index, including a final speed danger threshold TV3 and a final speed danger time threshold T3.
[0019] The final speed warning module determines the relationship between the average speed V of the fish and the final speed judgment index to obtain the water quality judgment result S3 and transmits it to the two-parameter linkage warning module.
[0020] The two-parameter linkage warning module processes the water quality judgment results S1, S2, and S3 in pairs to generate water quality judgment results S4, S5, and S6, and transmits them to the multi-parameter linkage warning module.
[0021] The multi-parameter linkage warning module comprehensively processes the water quality judgment results S4, S5, and S6 to obtain the final water quality judgment result.
[0022] Further, the step of determining the relationship between the average speed V of the fish and the initial speed judgment index to obtain the water quality judgment result S1 includes:
[0023] a1) judging whether the average speed of fish V < TV1 is true, if yes, the water quality judgment result S1 is normal water quality, otherwise, entering step a2);
[0024] a2) judging whether V < TV2 is true, if yes, entering step a3), otherwise, jumping to step a4);
[0025] a3) judging whether the duration of V < TV2 from the current frame exceeds the initial speed abnormal time threshold T1, if yes, the water quality judgment result S1 is abnormal water quality, otherwise, the water quality judgment result S1 is normal water quality;
[0026] a4) judging whether the duration of V > TV2 from the current frame exceeds the initial speed danger time threshold T2, if yes, the water quality judgment result S1 is dangerous water quality, otherwise, the water quality judgment result S1 is normal water quality.
[0027] Further, the step of the avoidance rate warning module judging the relationship between the fish avoidance rate and the avoidance rate judgment index to obtain the water quality judgment result S2 includes:
[0028] b1) initializing the fish school avoidance rate danger number tn2 = 0 and the fish school avoidance rate abnormal number tn1 = 0;
[0029] b2) calculating the fish avoidance rate R and judging whether the avoidance rate R > the avoidance rate danger threshold TR2 is true, if yes, setting the fish school avoidance rate danger number tn2 = tn2 + 1 and jumping to step b4); otherwise, entering step b3);
[0030] b3) judging whether the avoidance rate R > the avoidance rate abnormal threshold TR1 is true, if yes, setting the fish school avoidance rate abnormal number tn1 = tn1 + 1 and entering step b4); otherwise, directly entering step 4);
[0031] b4) judging whether the detection time reaches the warning judgment period T5, if yes, entering step 5), otherwise, returning to step b2);
[0032] b5) judging whether the fish school avoidance rate danger number tn2 > the danger number threshold TRN2 is true, if yes, the water quality judgment result S2 is dangerous water quality, otherwise, the water quality judgment result S2 is normal water quality;
[0033] judging whether the fish school avoidance rate abnormal number tn1 > the abnormal number threshold TRN1 is true, if yes, the water quality judgment result S2 is abnormal water quality, otherwise, the water quality judgment result S2 is normal water quality.
[0034] Further, the step of generating the water quality judgment result S3 by the end speed warning module comprises:
[0035] c1) whether the fish average speed V > TV3 is established, if yes, the water quality judgment result S3 is water quality normal, otherwise, step c2) is entered;
[0036] c2) whether the duration of the fish average speed V < TV3 from the current frame is more than the end speed abnormal time threshold T3 is established, if yes, the water quality judgment result S3 is water quality danger, otherwise, the water quality judgment result S3 is water quality normal.
[0037] Further, the step of generating the water quality judgment result S4 comprises:
[0038] d1) whether the water quality judgment result S1 is not water quality normal and the water quality judgment result S2 is water quality danger is established, if yes, the water quality judgment result S4 is water quality extremely dangerous, otherwise, step d2) is entered;
[0039] d2) whether the water quality judgment result S1 or the water quality judgment result S2 is water quality danger is established, if yes, the water quality judgment result S4 is water quality danger, otherwise, step d3) is entered;
[0040] d3) whether the water quality judgment result S1 and the water quality judgment result S2 are both water quality abnormal is established, if yes, the water quality judgment result S4 is water quality danger, otherwise, step d4) is entered;
[0041] d4) whether the water quality judgment result S1 or the water quality judgment result S2 is water quality abnormal is established, if yes, the water quality judgment result S4 is water quality abnormal, otherwise, the water quality judgment result S4 is water quality normal.
[0042] Further, the step of generating the water quality judgment result S5 comprises:
[0043] e1) whether the water quality judgment result S2 and the water quality judgment result S3 are both water quality danger is established, if yes, the water quality judgment result S5 is water quality extremely dangerous, otherwise, step e2) is entered;
[0044] e2) whether the water quality judgment result S2 or the water quality judgment result S3 is water quality danger is established, if yes, the water quality judgment result S5 is water quality danger, otherwise, step e3) is entered;
[0045] e3) whether the water quality judgment result S2 or the water quality judgment result S3 is water quality abnormal is established, if yes, the water quality judgment result S5 is water quality abnormal, otherwise, the water quality judgment result S5 is water quality normal.
[0046] Further, the step of generating the water quality judgment result S6 comprises:
[0047] f1) judging whether the water quality judgment result S1 is not water quality normal and the water quality judgment result S3 is water quality dangerous or not, if yes, the water quality judgment result S6 is water quality extremely dangerous, otherwise, entering step f2);
[0048] f2) judging whether the water quality judgment result S1 or the water quality judgment result S3 is water quality dangerous or not, if yes, the water quality judgment result S6 is water quality dangerous, otherwise, entering step f3);
[0049] f3) judging whether the water quality judgment result S1 and the water quality judgment result S3 are both water quality abnormal or not, if yes, the water quality judgment result S6 is water quality dangerous, otherwise, entering step f4);
[0050] f4) judging whether the water quality judgment result S1 or the water quality judgment result S3 is water quality abnormal or not, if yes, the water quality judgment result S6 is water quality abnormal, otherwise, the water quality judgment result S6 is water quality normal.
[0051] Further, the step of comprehensively processing the water quality judgment result S4, the water quality judgment result S5 and the water quality judgment result S6 to obtain the final water quality judgment result comprises:
[0052] g1) judging whether the water quality judgment result S4, the water quality judgment result S5 or the water quality judgment result S6 is water quality extremely dangerous or not, if yes, the final water quality judgment result is water quality extremely dangerous, otherwise, entering step g2);
[0053] g2) judging whether the water quality judgment result S4, the water quality judgment result S5 or the water quality judgment result S6 is water quality dangerous or not, if yes, the final water quality judgment result is water quality dangerous, otherwise, entering step 3);
[0054] g3) judging whether the water quality judgment result S4, the water quality judgment result S5 or the water quality judgment result S6 is water quality abnormal or not, if yes, the final water quality judgment result is water quality abnormal, otherwise, the final water quality judgment result is water quality normal.
[0055] Further, the method further comprises an alarm module;
[0056] When the final water quality judgment result is water quality abnormal, water quality dangerous or water quality extremely dangerous, the alarm module issues different degrees of alarm.
[0057] The technical effect of the present application is self-evident. The present application introduces the initial speed, the final speed and the avoidance rate of fish as water quality monitoring parameters, and jointly issues a warning. The highest priority warning state among the three joint warning results is selected as the system water quality monitoring result, thereby improving the water quality monitoring accuracy.
[0058] Deep analysis is conducted on each behavior parameter, the internal relation between the parameters is fully considered, multi-parameter verification and linkage early warning are conducted, the early warning strategies under the threshold of each behavior parameter are organically fused, the influence of interference factors can be effectively eliminated, and the accuracy of water quality monitoring system early warning can be improved and the false alarm rate can be reduced.
[0059] The application can effectively monitor water quality safety and reduce false alarm rate, has great help for improving water quality environment and improving water quality of people, in addition, the application can also effectively monitor various drinking water poisoning and other malignant events, and effectively protect human life and property safety. BRIEF DESCRIPTION OF DRAWINGS
[0060] Figure 1 To avoid the single parameter early warning strategy of the average speed of fish outside the area,
[0061] Figure 2 To avoid the single parameter early warning strategy of the rate,
[0062] Figure 3 To avoid the linkage early warning strategy of the initial speed and the rate,
[0063] Figure 4 To avoid the linkage early warning strategy of the rate and the final speed,
[0064] Figure 5 To avoid the linkage early warning strategy of the initial speed and the final speed,
[0065] Figure 6 To avoid the final linkage early warning strategy. DETAILED DESCRIPTION
[0066] The application will be further described below in combination with examples, but should not be understood as the above-mentioned subject matter of the application being limited to the following examples. According to ordinary technical knowledge and conventional means in the art, various substitutions and changes can be made without departing from the above-mentioned technical idea of the application, and all should be included in the protection scope of the application.
[0067] Example 1:
[0068] Referring to Figures 1 to 6 , the water quality safety multi-parameter linkage early warning system based on ichthyology includes a water tank, a camera, an initial speed early warning module, a final speed early warning module, a rate early warning module, a two-parameter linkage early warning module and a multi-parameter linkage early warning module.
[0069] The water tank is used for group feeding fish.
[0070] The water tank comprises a fish group activity area and an avoidance area; the water in the fish group activity area is a detection water sample; and the water in the avoidance area is a normal water sample.
[0071] The camera shoots fish activity images and transmits them to the initial speed early warning module, the final speed early warning module and the avoidance rate early warning module respectively;
[0072] The initial speed early warning module stores initial speed judgment indexes, including an initial speed abnormal threshold TV1, an initial speed danger threshold TV2, an initial speed abnormal time threshold T1 and an initial speed danger time threshold T2;
[0073] The initial speed early warning module processes the fish activity images to obtain the average speed V of the fish and transmits it to the final speed early warning module;
[0074] The initial speed early warning module judges the relationship between the average speed V of the fish and the initial speed judgment indexes to obtain the water quality judgment result S1 and transmits it to the two-parameter linkage early warning module;
[0075] The avoidance rate early warning module stores avoidance rate detection period T4, avoidance rate determination period T5 and avoidance rate judgment indexes;
[0076] The avoidance rate judgment indexes include an avoidance rate abnormal threshold TR1, a danger threshold TR2, an abnormal number threshold TRN1 within the avoidance rate determination period T5 and a danger number threshold TRN2 within the avoidance rate determination period T5;
[0077] The avoidance rate early warning module processes the fish activity images according to the avoidance rate detection period T4 to obtain the avoidance rate R of the fish;
[0078] The avoidance rate early warning module judges the relationship between the avoidance rate R of the fish and the avoidance rate judgment indexes to obtain the water quality judgment result S2 and transmits it to the two-parameter linkage early warning module;
[0079] The final speed early warning module stores final speed judgment indexes, including a final speed danger threshold TV3 and a final speed danger time threshold T3;
[0080] The final speed early warning module judges the relationship between the average speed V of the fish and the final speed judgment indexes to obtain the water quality judgment result S3 and transmits it to the two-parameter linkage early warning module;
[0081] The two-parameter linkage early warning module processes the water quality judgment results S1, S2 and S3 in pairs to generate water quality judgment results S4, S5 and S6 and transmits them to the multi-parameter linkage early warning module;
[0082] The multi-parameter linkage early warning module comprehensively processes the water quality judgment results S4, S5 and S6 to obtain the final water quality judgment result.
[0083] The initial speed warning module judges the relationship between the average speed V of fish and the initial speed judgment index to obtain the water quality judgment result S1, and the steps include:
[0084] a1) judging whether the average speed V of fish is less than TV1, if yes, the water quality judgment result S1 is normal water quality, otherwise, step a2) is entered;
[0085] a2) judging whether V is less than TV2, if yes, step a3) is entered, otherwise, step a4) is jumped to;
[0086] a3) judging whether the duration of the average speed V of fish being less than TV2 from the current frame exceeds the initial speed abnormal time threshold T1, if yes, the water quality judgment result S1 is abnormal water quality, otherwise, the water quality judgment result S1 is normal water quality;
[0087] a4) judging whether the duration of the average speed V of fish being greater than TV2 from the current frame exceeds the initial speed danger time threshold T2, if yes, the water quality judgment result S1 is dangerous water quality, otherwise, the water quality judgment result S1 is normal water quality.
[0088] The avoidance rate warning module judges the relationship between the avoidance rate of fish and the avoidance rate judgment index to obtain the water quality judgment result S2, and the steps include:
[0089] b1) initializing the fish school avoidance rate danger number tn2 = 0 and the fish school avoidance rate abnormal number tn1 = 0;
[0090] b2) calculating the avoidance rate R of fish, and judging whether the avoidance rate R is greater than the avoidance rate danger threshold TR2, if yes, the fish school avoidance rate danger number tn2 = tn2 + 1 is set, and step b4) is jumped to; otherwise, step b3) is entered;
[0091] b3) judging whether the avoidance rate R is greater than the avoidance rate abnormal threshold TR1, if yes, the fish school avoidance rate abnormal number tn1 = tn1 + 1 is set, and step b4) is entered; otherwise, step 4) is directly entered;
[0092] b4) judging whether the detection time reaches the warning judgment period T5, if yes, step 5) is entered, otherwise, step b2) is returned;
[0093] b5) judging whether the fish school avoidance rate danger number tn2 is greater than the danger number threshold TRN2, if yes, the water quality judgment result S2 is dangerous water quality, otherwise, the water quality judgment result S2 is normal water quality;
[0094] judging whether the fish school avoidance rate abnormal number tn1 is greater than the abnormal number threshold TRN1, if yes, the water quality judgment result S2 is abnormal water quality, otherwise, the water quality judgment result S2 is normal water quality.
[0095] The step of determining the water quality judgment result S3 by the terminal speed early warning module judging the relationship between the average speed V of the fish and the terminal speed judgment index comprises:
[0096] c1) whether the average speed V of the fish is greater than TV3, if yes, the water quality judgment result S3 is normal water quality, otherwise, step c2) is entered;
[0097] c2) whether the duration of the average speed V of the fish being less than TV3 from the current frame exceeds a terminal speed abnormal time threshold T3, if yes, the water quality judgment result S3 is dangerous (or abnormal) water quality, otherwise, the water quality judgment result S3 is normal water quality.
[0098] The step of generating the water quality judgment result S4 comprises:
[0099] d1) whether the water quality judgment result S1 is not normal water quality and the water quality judgment result S2 is dangerous water quality, if yes, the water quality judgment result S4 is extremely dangerous water quality, otherwise, step d2) is entered;
[0100] d2) whether the water quality judgment result S1 or the water quality judgment result S2 is dangerous water quality, if yes, the water quality judgment result S4 is dangerous water quality, otherwise, step d3) is entered;
[0101] d3) whether the water quality judgment result S1 and the water quality judgment result S2 are both abnormal water quality, if yes, the water quality judgment result S4 is dangerous water quality, otherwise, step d4) is entered;
[0102] d4) whether the water quality judgment result S1 or the water quality judgment result S2 is abnormal water quality, if yes, the water quality judgment result S4 is abnormal water quality, otherwise, the water quality judgment result S4 is normal water quality.
[0103] The step of generating the water quality judgment result S5 comprises:
[0104] e1) whether the water quality judgment result S2 and the water quality judgment result S3 are both dangerous water quality, if yes, the water quality judgment result S5 is extremely dangerous water quality, otherwise, step e2) is entered;
[0105] e2) whether the water quality judgment result S2 or the water quality judgment result S3 is dangerous water quality, if yes, the water quality judgment result S5 is dangerous water quality, otherwise, step e3) is entered;
[0106] e3) whether the water quality judgment result S2 or the water quality judgment result S3 is abnormal water quality, if yes, the water quality judgment result S5 is abnormal water quality, otherwise, the water quality judgment result S5 is normal water quality.
[0107] The step of generating the water quality judgment result S6 includes:
[0108] f1) judging whether the water quality judgment result S1 is not normal and the water quality judgment result S3 is dangerous, if yes, the water quality judgment result S6 is extremely dangerous, otherwise, entering step f2);
[0109] f2) judging whether the water quality judgment result S1 or the water quality judgment result S3 is dangerous, if yes, the water quality judgment result S6 is dangerous, otherwise, entering step f3);
[0110] f3) judging whether the water quality judgment result S1 and the water quality judgment result S3 are abnormal, if yes, the water quality judgment result S6 is dangerous, otherwise, entering step f4);
[0111] f4) judging whether the water quality judgment result S1 or the water quality judgment result S3 is abnormal, if yes, the water quality judgment result S6 is abnormal, otherwise, the water quality judgment result S6 is normal.
[0112] The step of comprehensively processing the water quality judgment result S4, the water quality judgment result S5 and the water quality judgment result S6 to obtain a final water quality judgment result includes:
[0113] g1) judging whether the water quality judgment result S4, the water quality judgment result S5 or the water quality judgment result S6 is extremely dangerous, if yes, the final water quality judgment result is extremely dangerous, otherwise, entering step g2);
[0114] g2) judging whether the water quality judgment result S4, the water quality judgment result S5 or the water quality judgment result S6 is dangerous, if yes, the final water quality judgment result is dangerous, otherwise, entering step 3);
[0115] g3) judging whether the water quality judgment result S4, the water quality judgment result S5 or the water quality judgment result S6 is abnormal, if yes, the final water quality judgment result is abnormal, otherwise, the final water quality judgment result is normal.
[0116] The system further includes an alarm module;
[0117] When the final water quality judgment result is abnormal, dangerous or extremely dangerous, the alarm module issues different degrees of alarm.
[0118] Embodiment 2:
[0119] Referring to Figures 1 to 6 , a water quality safety multi-parameter linkage early warning system based on ichthyology, the content is as follows:
[0120] 1. Speed-based water quality early warning strategy
[0121] The speed-based water quality early warning strategy process is shown in Figure 1 The specific early warning strategy is as follows: when the fish school is early exposed to a high concentration of pollutants, due to the intense movement caused by the stress reaction, the speed of the fish school is too fast, the initial speed anomaly threshold TV1 and the danger threshold TV2 can be determined; when the fish school is exposed to a low concentration or a less toxic pollutant environment, it is close to death after a long time, resulting in slow speed of the fish school, the final speed danger threshold TV3 can be determined; in addition, the time threshold in each state needs to be determined, and the current state can be determined only when the fish school speed reaches the speed threshold and the duration reaches the time threshold, that is, the initial speed anomaly time threshold T1, the speed danger time threshold T2 and the final speed danger time threshold T3 are determined, which respectively represent the time warning line that needs to be maintained after the speed exceeds the corresponding threshold.
[0122] After determining the above thresholds, the average speed V of the fish school outside the avoidance area is calculated according to the fish school detection and tracking results, and the relationship between V and TV1, TV2 and TV3 is judged. If V is less than TV3, it is judged whether the duration of speed V less than TV3 from the current frame exceeds T3, if it exceeds, the system issues an extremely dangerous water quality warning at this time, otherwise it indicates that the fish school behavior is normal; if V is greater than TV1 and less than TV2, it is judged whether the duration of speed V greater than TV1 and less than TV2 from the current frame exceeds the abnormal time threshold T1, if it exceeds, the system issues an abnormal water quality warning at this time, otherwise it indicates that the fish school behavior is normal; if V is greater than TV2, it is judged whether the duration of V greater than TV2 from the current frame exceeds T2, if it exceeds, the system issues a dangerous water quality warning at this time, otherwise it indicates that the fish school behavior is normal; finally, when V is greater than TV3 and less than TV1, it indicates that the speed of the fish school is within the normal range at this time, the water quality is normal, and the system does not issue any warning.
[0123] 2. Water quality early warning strategy based on avoidance rate
[0124] The coordinates of the avoidance area in the fish tank are divided, and the fish tank image collected by the camera is intercepted. The coordinates of the avoidance area are determined by manually or automatically framing the image. Then, it is judged whether the centroid coordinates of each fish are within the coordinate range of the avoidance area, if within the range, the avoidance times are added by 1, if not, the next time is re-judged, and finally the avoidance times within a certain time are counted to determine the value of the avoidance rate.
[0125] The water quality early warning strategy based on avoidance rate process is shown in Figure 2The specific early warning strategy is as follows: first, the avoidance rate detection period T4 is set, the known avoidance rate refers to the number of times that the fish population enters the avoidance area within T4, based on which the avoidance rate anomaly threshold TR1 and the danger threshold TR2 within T4 are set, which respectively represent the warning values of the number of times that the fish population enters the avoidance area within the avoidance rate detection period T4 reaching the abnormal and dangerous behavior state; then the avoidance rate determination period T5 is set, which represents the time period for single water quality state analysis using the avoidance rate parameter, based on which the abnormal number threshold TRN1 and the dangerous number threshold TRN2 within T5 are set, which respectively represent the warning values of the cumulative abnormal behavior number and the cumulative dangerous behavior number of the fish population within the early warning determination period T5, and when the values are exceeded, the system will produce corresponding alarms.
[0126] After the above parameters are determined, the average avoidance rate R of the fish population is calculated according to the fish population detection and tracking results and T4, and the relationship between R and TR1 and TR2 is judged in real time: if R is greater than TR2, the fish population behavior dangerous number tn2 increases by 1, if R is greater than TR1, the fish population behavior abnormal number tn1 increases by 1, otherwise, both remain unchanged. At this time, one avoidance rate detection period is completed, and it is further judged whether the water quality state detection time reaches T5, if yes, the current T5 fish population avoidance rate abnormal number tn1 and avoidance rate dangerous number tn2 are recorded, and the relationship between tn1, tn2 and TRN1, TRN2 is judged to analyze the water quality state at this time: when tn1 is greater than TRN1, the system issues a water quality abnormal alarm; when tn2 is greater than TRN2, the system issues a water quality danger alarm; otherwise, it indicates that the fish population behavior parameter is normal, and the system does not alarm; if T5 is not reached at this time, the above process is repeated in the next avoidance rate detection period until the early warning determination period T5 is reached. Finally, the water quality state analysis is completed, and the fish population abnormal number counter, the dangerous number counter and the detection time are set to zero, so as to continue to analyze the next early warning determination period water quality early warning state.
[0127] The water quality early warning strategy based on the fish population avoidance rate parameter is mainly formulated according to the instinctive escape from poison of the fish population when exposed to pollutants.
[0128] 3. Combination of two parameters
[0129] According to the relevant experiments, when the fish population is exposed to toxicants, the fish population appears to accelerate to avoid the area to escape, at this time the fish population speed will increase before the avoidance rate, the initial speed and the avoidance rate linkage early warning strategy is proposed, the specific process is as follows: extract the single parameter warning results S1 and S2 of initial speed, avoidance rate, and carry out joint analysis to judge the current water quality warning state. When S1 first appears abnormal state and then S2 appears dangerous or S1 and S2 both appear dangerous state, it indicates that the fish population appears to accelerate to avoid the area to escape, at this time the system issues water quality extremely dangerous alarm; When excluding the above conditions, one of S1 and S2 with the highest priority state is dangerous or both are abnormal, at this time the system issues water quality dangerous alarm; When one of S1 and S2 with the highest priority state is abnormal, the system issues water quality abnormal alarm; Finally, when S1 and S2 are both normal, the water quality parameter is normal.
[0130] When the fish population is exposed to low concentration or weak toxicity pollutants environment, it will be close to death after a long time, resulting in slow fish population speed, the avoidance rate and the end speed linkage early warning strategy is proposed, the specific process is as follows: extract the single parameter warning results S2 and S3 of avoidance rate and end speed, and carry out joint analysis to judge the current water quality warning state. Only when S2 and S3 are both dangerous state, the system issues water quality extremely dangerous alarm, otherwise the system warning state is the highest priority alarm state of the two, if both are normal state, the water quality parameter is normal.
[0131] According to the relevant experiments, when the fish population is exposed to strong toxicity or high concentration toxicants, the speed first increases rapidly, and then the fish body is affected by the toxicants, the fish population speed decreases greatly, the avoidance rate and the end speed linkage early warning strategy is proposed, the specific process is as follows: extract the single parameter warning results S1 and S3 of initial speed and end speed, and carry out joint analysis to judge the current water quality warning state. When S1 first appears abnormal state and then S3 appears dangerous or S1 and S3 both appear dangerous state, it indicates that the fish population behavior appears to increase speed first and then decrease, at this time the system issues water quality extremely dangerous alarm; When excluding the above conditions, one of S1 and S3 with the highest priority state is dangerous or both are abnormal, at this time the system issues water quality dangerous alarm; When one of S1 and S3 with the highest priority state is abnormal, the system issues water quality abnormal alarm; Finally, when S1 and S3 are both normal, the water quality parameter is normal.
[0132] 4、Multiple parameter combination
[0133] Since the above three linkage early warning methods are running in the system at the same time, the final water quality monitoring and early warning flow chart is shown in the above figure. Its main function is to select the highest priority warning state of the three linkage early warning results as the system water quality monitoring result.
[0134] Example 3:
[0135] The water quality safety multi-parameter linkage early warning system based on fish behavior includes a water tank, a camera, an initial speed early warning module, a terminal speed early warning module, an avoidance rate early warning module, a two-parameter linkage early warning module, and a multi-parameter linkage early warning module. The water tank is used for group feeding fish.
[0136] The water tank includes a fish group activity area and an avoidance area. The water in the fish group activity area is a detection water sample, and the water in the avoidance area is a normal water sample.
[0137] The camera captures fish activity images and transmits them to the initial speed early warning module, the terminal speed early warning module, and the avoidance rate early warning module, respectively.
[0138] The initial speed early warning module stores initial speed judgment indexes, including an initial speed abnormal threshold TV1, an initial speed danger threshold TV2, an initial speed abnormal time threshold T1, and an initial speed danger time threshold T2.
[0139] The initial speed early warning module processes fish activity images to obtain an average fish speed V and transmits it to the terminal speed early warning module.
[0140] The initial speed early warning module determines the relationship between the average fish speed V and the initial speed judgment indexes to obtain a water quality judgment result S1 and transmits it to the two-parameter linkage early warning module.
[0141] The avoidance rate early warning module stores avoidance rate detection periods T4, avoidance rate determination periods T5, and avoidance rate judgment indexes.
[0142] The avoidance rate judgment indexes include an avoidance rate abnormal threshold TR1, a danger threshold TR2, an abnormal number threshold TRN1 within the avoidance rate determination period T5, and a danger number threshold TRN2 within the avoidance rate determination period T5.
[0143] The avoidance rate early warning module processes fish activity images according to the avoidance rate detection period T4 to obtain a fish avoidance rate R.
[0144] The avoidance rate early warning module determines the relationship between the fish avoidance rate R and the avoidance rate judgment indexes to obtain a water quality judgment result S2 and transmits it to the two-parameter linkage early warning module.
[0145] The terminal speed early warning module stores terminal speed judgment indexes, including a terminal speed danger threshold TV3 and a terminal speed danger time threshold T3.
[0146] The terminal speed early warning module determines the relationship between the average fish speed V and the terminal speed judgment indexes to obtain a water quality judgment result S3 and transmits it to the two-parameter linkage early warning module.
[0147] The two-parameter linkage early warning module processes the water quality judgment results S1, S2, and S3 in pairs to generate water quality judgment results S4, S5, and S6, and transmits them to the multi-parameter linkage early warning module;
[0148] The multi-parameter linkage early warning module comprehensively processes the water quality judgment results S4, S5, and S6 to obtain the final water quality judgment result.
[0149] Embodiment 4:
[0150] The water quality safety multi-parameter linkage early warning system based on fish behavior mainly includes the contents of Embodiment 3, wherein the step of determining the relationship between the average fish speed V and the initial speed judgment index to obtain the water quality judgment result S1 includes:
[0151] 1) Determine whether the average fish speed V < TV1 is true, if yes, the water quality judgment result S1 is normal, otherwise, go to step 2);
[0152] 2) Determine whether V < TV2 is true, if yes, go to step 3), otherwise, jump to step 4);
[0153] 3) Determine whether the duration of the average fish speed V < TV2 from the current frame exceeds the initial speed abnormal time threshold T1, if yes, the water quality judgment result S1 is abnormal, otherwise, the water quality judgment result S1 is normal;
[0154] 4) Determine whether the duration of the average fish speed V > TV2 from the current frame exceeds the initial speed danger time threshold T2, if yes, the water quality judgment result S1 is dangerous, otherwise, the water quality judgment result S1 is normal.
[0155] Embodiment 5:
[0156] The water quality safety multi-parameter linkage early warning system based on fish behavior mainly includes the contents of Embodiment 3, wherein the step of determining the relationship between the fish avoidance rate and the avoidance rate judgment index to obtain the water quality judgment result S2 includes:
[0157] 1) Initialize the fish avoidance rate danger count tn2 = 0 and the fish avoidance rate abnormal count tn1 = 0;
[0158] 2) Calculate the fish avoidance rate R and determine whether the avoidance rate R > avoidance rate danger threshold TR2 is true, if yes, set the fish avoidance rate danger count tn2 = tn2 + 1 and jump to step 4); otherwise, go to step 3);
[0159] 3) judge whether the avoidance rate R is greater than the avoidance rate abnormal threshold TR1, if yes, let the fish school avoidance rate abnormal times tn1 = tn1 + 1, and enter step 4); otherwise, directly enter step 4);
[0160] 4) judge whether the detection time reaches the early warning judgment period T5, if yes, enter step 5), otherwise, return to step 2);
[0161] 5) judge whether the fish school avoidance rate danger times tn2 is greater than the danger times threshold TRN2, if yes, the water quality judgment result S2 is water quality danger, otherwise, the water quality judgment result S2 is water quality normal;
[0162] judge whether the fish school avoidance rate abnormal times tn1 is greater than the abnormal times threshold TRN1, if yes, the water quality judgment result S2 is water quality abnormal, otherwise, the water quality judgment result S2 is water quality normal.
[0163] Example 6:
[0164] The water quality safety multi-parameter linkage early warning system based on ichthyology mainly includes the contents of example 3, wherein the step of judging the relationship between the average speed V of fish and the terminal speed judgment index of the terminal speed early warning module to obtain the water quality judgment result S3 includes:
[0165] 1) whether the average speed V of fish is greater than TV3, if yes, the water quality judgment result S3 is water quality normal, otherwise, enter step 2);
[0166] 2) judge whether the duration of the average speed V of fish being less than TV3 from the current frame exceeds the terminal speed abnormal time threshold T3, if yes, the water quality judgment result S3 is water quality danger, otherwise, the water quality judgment result S3 is water quality normal.
[0167] Example 7:
[0168] The water quality safety multi-parameter linkage early warning system based on ichthyology mainly includes the contents of example 3, wherein the step of generating the water quality judgment result S4 includes:
[0169] 1) judge whether the water quality judgment result S1 is not water quality normal and the water quality judgment result S2 is water quality danger, if yes, the water quality judgment result S4 is water quality extremely dangerous, otherwise, enter step 2);
[0170] 2) judge whether the water quality judgment result S1 or the water quality judgment result S2 is water quality danger, if yes, the water quality judgment result S4 is water quality danger, otherwise, enter step 3);
[0171] 3) judging whether the water quality judging result S1 and the water quality judging result S2 are both water quality abnormality, if yes, the water quality judging result S4 is water quality danger, otherwise, entering step 4);
[0172] 4) judging whether the water quality judging result S1 or the water quality judging result S2 is water quality abnormality, if yes, the water quality judging result S4 is water quality abnormality, otherwise, the water quality judging result S4 is water quality normal.
[0173] Embodiment 8:
[0174] The water quality safety multi-parameter linkage early warning system based on fish behavior is mainly shown in embodiment 3, wherein the step of generating the water quality judging result S5 comprises:
[0175] 1) judging whether the water quality judging result S2 and the water quality judging result S3 are both water quality danger, if yes, the water quality judging result S5 is water quality extremely dangerous, otherwise, entering step 2);
[0176] 2) judging whether the water quality judging result S2 or the water quality judging result S3 is water quality danger, if yes, the water quality judging result S5 is water quality danger, otherwise, entering step 3);
[0177] 3) judging whether the water quality judging result S2 or the water quality judging result S3 is water quality abnormality, if yes, the water quality judging result S5 is water quality abnormality, otherwise, the water quality judging result S5 is water quality normal.
[0178] Embodiment 9:
[0179] The water quality safety multi-parameter linkage early warning system based on fish behavior is mainly shown in embodiment 3, wherein the step of generating the water quality judging result S6 comprises:
[0180] 1) judging whether the water quality judging result S1 is not water quality normal and the water quality judging result S3 is water quality danger, if yes, the water quality judging result S6 is water quality extremely dangerous, otherwise, entering step 2);
[0181] 2) judging whether the water quality judging result S1 or the water quality judging result S3 is water quality danger, if yes, the water quality judging result S6 is water quality danger, otherwise, entering step 3);
[0182] 3) judging whether the water quality judging result S1 and the water quality judging result S3 are both water quality abnormality, if yes, the water quality judging result S6 is water quality danger, otherwise, entering step 4);
[0183] 4) judging whether the water quality judging result S1 or the water quality judging result S3 is water quality abnormality, if yes, the water quality judging result S6 is water quality abnormality, otherwise, the water quality judging result S6 is water quality normal.
[0184] Example 10:
[0185] The water quality safety multi-parameter linkage early warning system based on ichthyology mainly includes the contents of Example 3, wherein the steps of comprehensively processing the water quality judgment result S4, the water quality judgment result S5 and the water quality judgment result S6 to obtain the final water quality judgment result include:
[0186] 1) judging whether the water quality judgment result S4, the water quality judgment result S5 or the water quality judgment result S6 is water quality extremely dangerous, if yes, the final water quality judgment result is water quality extremely dangerous, otherwise, step 2) is entered;
[0187] 2) judging whether the water quality judgment result S4, the water quality judgment result S5 or the water quality judgment result S6 is water quality dangerous, if yes, the final water quality judgment result is water quality dangerous, otherwise, step 3) is entered;
[0188] 3) judging whether the water quality judgment result S4, the water quality judgment result S5 or the water quality judgment result S6 is water quality abnormal, if yes, the final water quality judgment result is water quality abnormal, otherwise, the final water quality judgment result is water quality normal.
[0189] Example 11:
[0190] The water quality safety multi-parameter linkage early warning system based on ichthyology mainly includes the contents of Example 3, wherein, the system further comprises an alarm module;
[0191] When the final water quality judgment result is water quality abnormal, water quality dangerous or water quality extremely dangerous, the alarm module sends different degrees of alarm.
Claims
1. A water quality safety multi-parameter linkage early warning system based on fish behavior, characterized in that: The system comprises a water tank, a camera, an initial speed warning module, a final speed warning module, an avoidance rate warning module, a two-parameter linkage warning module and a multi-parameter linkage warning module. The water tank comprises a fish group activity area and an avoidance area. The camera captures fish activity images and transmits them to the initial speed warning module, the final speed warning module and the avoidance rate warning module. The initial speed warning module stores initial speed judgment indexes, including an initial speed abnormal threshold TV1, an initial speed danger threshold TV2, an initial speed abnormal time threshold T1 and an initial speed danger time threshold T2. The initial speed warning module processes fish activity images to obtain an average fish speed V and transmits it to the final speed warning module. The initial speed warning module determines the relationship between the average fish speed V and the initial speed judgment indexes to obtain a water quality judgment result S1 and transmits it to the two-parameter linkage warning module. The avoidance rate warning module stores avoidance rate detection and judgment indexes, including an avoidance rate abnormal threshold TR1, a danger threshold TR2, an abnormal number threshold TRN1 within the avoidance rate detection period T4 and a danger number threshold TRN2 within the avoidance rate judgment period T5. The avoidance rate warning module processes fish activity images according to the avoidance rate detection period T4 to obtain an avoidance rate R of the fish. The avoidance rate warning module determines the relationship between the avoidance rate R of the fish and the avoidance rate judgment indexes to obtain a water quality judgment result S2 and transmits it to the two-parameter linkage warning module. The final speed warning module stores final speed judgment indexes, including a final speed danger threshold TV3 and a final speed danger time threshold T3. The final speed warning module determines the relationship between the average fish speed V and the final speed judgment indexes to obtain a water quality judgment result S3 and transmits it to the two-parameter linkage warning module. The two-parameter linkage warning module processes the water quality judgment results S1, S2 and S3 in pairs to generate water quality judgment results S4, S5 and S6 and transmits them to the multi-parameter linkage warning module. The multi-parameter linkage warning module comprehensively processes the water quality judgment results S4, S5 and S6 to obtain a final water quality judgment result. The step of determining the relationship between the average fish speed V and the initial speed judgment indexes to obtain the water quality judgment result S1 comprises:
2. The fish behavior-based water quality safety multi-parameter linkage early warning system according to claim 1, characterized in that, 1) determining whether the average fish speed V < TV1 is true, if yes, the water quality judgment result S1 is normal, otherwise, step 2) is entered; 2) determining whether V < TV2 is true, if yes, step 3) is entered, otherwise, step 4) is jumped to; 3) determining whether the average fish speed V < TV2 is true, if yes, the water quality judgment result S1 is abnormal, otherwise, step 4) is entered; 4) determining whether the average fish speed V < TV2 is true, if yes, the water quality judgment result S1 is dangerous, otherwise, the water quality judgment result S1 is normal. 3) judging whether the duration of fish average speed V < TV2 from the current frame exceeds the initial speed abnormal time threshold T1, if yes, the water quality judgment result S1 is water quality abnormal, otherwise, the water quality judgment result S1 is water quality normal; 4) judging whether the duration of fish average speed V > TV2 from the current frame exceeds the initial speed danger time threshold T2, if yes, the water quality judgment result S1 is water quality danger, otherwise, the water quality judgment result S1 is water quality normal.
3. The fish behavior-based water quality safety multi-parameter linkage early warning system according to claim 1, characterized in that, The steps of obtaining the water quality judgment result S2 by the avoidance rate early warning module judging the relationship between the fish avoidance rate and the avoidance rate judgment index include: 1) initializing the fish school avoidance rate danger number tn2 = 0 and the fish school avoidance rate abnormal number tn1 = 0; 2) calculating the fish avoidance rate R and judging whether the avoidance rate R > avoidance rate danger threshold TR2 is true, if yes, setting the fish school avoidance rate danger number tn2 = tn2 + 1 and jumping to step 4), otherwise, entering step 3); 3) judging whether the avoidance rate R > avoidance rate abnormal threshold TR1 is true, if yes, setting the fish school avoidance rate abnormal number tn1 = tn1 + 1 and entering step 4), otherwise, directly entering step 4); 4) judging whether the detection time reaches the early warning judgment period T5, if yes, entering step 5), otherwise, returning to step 2); 5) judging whether the fish school avoidance rate danger number tn2 > danger number threshold TRN2 is true, if yes, the water quality judgment result S2 is water quality danger, otherwise, the water quality judgment result S2 is water quality normal; judging whether the fish school avoidance rate abnormal number tn1 > abnormal number threshold TRN1 is true, if yes, the water quality judgment result S2 is water quality abnormal, otherwise, the water quality judgment result S2 is water quality normal.
4. The fish behavior-based water quality safety multi-parameter linkage early warning system according to claim 1, characterized in that, The steps of obtaining the water quality judgment result S3 by the terminal speed early warning module judging the relationship between the fish average speed V and the terminal speed judgment index include: 1) judging whether the fish average speed V > TV3 is true, if yes, the water quality judgment result S3 is water quality normal, otherwise, entering step 2); 2) judging whether the duration of fish average speed V < TV3 from the current frame exceeds the terminal speed abnormal time threshold T3, if yes, the water quality judgment result S3 is water quality danger, otherwise, the water quality judgment result S3 is water quality normal.
5. The fish behavior-based water quality safety multi-parameter linkage early warning system according to claim 1, characterized in that, The steps of generating the water quality judgment result S4 include: 1) judging whether the water quality judgment result S1 is not water quality normal and the water quality judgment result S2 is water quality danger is true, if yes, the water quality judgment result S4 is water quality extremely dangerous, otherwise, entering step 2); 2) judging whether the water quality judgment result S1 or the water quality judgment result S2 is water quality danger is true, if yes, the water quality judgment result S4 is water quality danger, otherwise, entering step 3); 3) judging whether the water quality judgment result S1 and the water quality judgment result S2 are both water quality abnormal is true, if yes, the water quality judgment result S4 is water quality danger, otherwise, entering step 4); 4) judging whether the water quality judgment result S1 or the water quality judgment result S2 is water quality abnormal is true, if yes, the water quality judgment result S4 is water quality abnormal, otherwise, the water quality judgment result S4 is water quality normal.
6. The fish behavior-based water quality safety multi-parameter linkage early warning system according to claim 1, characterized in that, The step of generating the water quality judgment result S5 includes: 1) judging whether the water quality judgment result S2 and the water quality judgment result S3 are both water quality dangerous or not, if yes, the water quality judgment result S5 is water quality extremely dangerous, otherwise, entering step 2); 2) judging whether the water quality judgment result S2 or the water quality judgment result S3 is water quality dangerous or not, if yes, the water quality judgment result S5 is water quality dangerous, otherwise, entering step 3); 3) judging whether the water quality judgment result S2 or the water quality judgment result S3 is water quality abnormal or not, if yes, the water quality judgment result S5 is water quality abnormal, otherwise, the water quality judgment result S5 is water quality normal.
7. The fish behavior-based water quality safety multi-parameter linkage early warning system according to claim 1, characterized in that, The step of generating the water quality judgment result S6 includes: 1) judging whether the water quality judgment result S1 is not water quality normal and the water quality judgment result S3 is water quality dangerous or not, if yes, the water quality judgment result S6 is water quality extremely dangerous, otherwise, entering step 2); 2) judging whether the water quality judgment result S1 or the water quality judgment result S3 is water quality dangerous or not, if yes, the water quality judgment result S6 is water quality dangerous, otherwise, entering step 3); 3) judging whether the water quality judgment result S1 and the water quality judgment result S3 are both water quality abnormal or not, if yes, the water quality judgment result S6 is water quality dangerous, otherwise, entering step 4); 4) judging whether the water quality judgment result S1 or the water quality judgment result S3 is water quality abnormal or not, if yes, the water quality judgment result S6 is water quality abnormal, otherwise, the water quality judgment result S6 is water quality normal.
8. The fish behavior-based water quality safety multi-parameter linkage early warning system according to claim 1, characterized in that, The step of comprehensively processing the water quality judgment result S4, the water quality judgment result S5 and the water quality judgment result S6 to obtain the final water quality judgment result includes: 1) judging whether the water quality judgment result S4, the water quality judgment result S5 or the water quality judgment result S6 is water quality extremely dangerous or not, if yes, the final water quality judgment result is water quality extremely dangerous, otherwise, entering step 2); 2) judging whether the water quality judgment result S4, the water quality judgment result S5 or the water quality judgment result S6 is water quality dangerous or not, if yes, the final water quality judgment result is water quality dangerous, otherwise, entering step 3); 3) judging whether the water quality judgment result S4, the water quality judgment result S5 or the water quality judgment result S6 is water quality abnormal or not, if yes, the final water quality judgment result is water quality abnormal, otherwise, the final water quality judgment result is water quality normal.
9. The fish behavior-based water quality safety multi-parameter linkage early warning system according to claim 1, characterized in that: The alarm module is further included; When the final water quality judgment result is water quality abnormal, water quality dangerous or water quality extremely dangerous, the alarm module sends different degrees of alarm.